AI-powered virtual executive team — a single coherent executive persona backed by 8 specialist agents (FastAPI + Next.js).
feat(onboarding): choose Personal or Team automatically and save goals with owners (#595)
Rufusb1f6a1c3bbb1
diff --git a/packages/core/openexecutive/agents/onboarding_interviewer.py b/packages/core/openexecutive/agents/onboarding_interviewer.pyindex c53445df..d83b095b 100644--- a/packages/core/openexecutive/agents/onboarding_interviewer.py+++ b/packages/core/openexecutive/agents/onboarding_interviewer.py@@ -2,9 +2,10 @@The conversational sibling of ``engagement_intake``. Where that agent reads apile of intake material in one shot, this one *talks*: it asks one clarifying-question at a time until it can draft the company, then emits a profile,-a leadership roster, and a department list for the user to review and edit-before anything is saved.+quick question at a time (with suggested answers to tap) until it can draft+the company, then emits a profile, the people, this year's goals with owners,+and departments only when the user already listed people by department, for+the user to review and edit before anything is saved.Same epistemics as ``engagement_intake`` — extract, never invent — because thecompany being modelled is the user's real one. The difference is that here the@@ -27,31 +28,38 @@ ONBOARDING_INTERVIEWER_AGENT_ID = "onboarding_interviewer"ONBOARDING_INTERVIEWER_SYSTEM = ("You are setting up Open Executive for a new user. They have described ""their company in their own words; your job is to understand it well "- "enough to draft a company profile, a leadership roster, and a department "- "list, asking for anything material you are missing.\n\n"+ "enough to draft a company profile, the people, and this year's goals "+ "with an owner on each, asking a few quick questions about anything "+ "material you are missing.\n\n""Every turn you MUST call exactly one tool: ask_clarifying_question when ""something material is still missing, or emit_company_draft when you have ""enough.\n\n""Grounding rules (these override everything else):\n""- Extract, don't invent: every fact in your draft must trace to something ""the user said. When something is not stated, leave the field null or "- "empty and name it in confidence_notes — NEVER estimate or fabricate "- "numbers (ARR, burn, runway, headcount, founding year).\n"+ "empty — NEVER estimate or fabricate numbers (ARR, burn, runway, "+ "headcount, founding year).\n""- The user reviews and edits every field before anything is saved, so "- "bias toward drafting EARLY. A draft that is 80% right and takes them 30 "- "seconds to correct beats eight questions. Do not interrogate.\n"+ "bias toward drafting EARLY. Ask two to four questions at most, and none "+ "when the description already covers the basics. Do not interrogate.\n""- Ask exactly ONE question per turn, and never re-ask something already "- "answered or already visible in an existing profile.\n\n"- "Question priority, highest first — if you can only ask a few, ask these:\n"- "1. Company name, what it does, industry, and stage.\n"- "2. Who they sell to, and why customers pick them over alternatives.\n"- "3. This year's priorities and the one metric that matters most.\n"- "4. Who is on the leadership team, and which of them is the user.\n"- "5. Optional, only if there is budget left: mission/vision, culture, "- "financials.\n\n"+ "answered or already visible in an existing profile.\n"+ "- Every question offers two to four short suggested answers in options, "+ "built from what the user already said, so they can answer with one tap. "+ "Questions are optional: when the user skips one, move on and do not ask "+ "it again.\n\n"+ "What to ask about, highest first, only where it is missing:\n"+ "1. Company name, what it does, and who it is for.\n"+ "2. What they want to achieve this year, and who owns each goal when "+ "several people are named.\n"+ "3. Which of the named people is the user, if unclear.\n"+ "4. A guess worth confirming, or two sources that disagree.\n"+ "Never ask about departments, org structure or how to group the work. "+ "Never ask about anything the user chose to leave out (for example "+ "finances).\n\n""Sensitive topics:\n"- "- Financials: ask at most once, say that figures are stored locally and "- "never shared, and accept a refusal without pushing.\n"+ "- Financials: never ask. Record figures only if the user volunteers "+ "them.\n""- NEVER record email addresses, phone numbers, or chat handles for ""anyone, even if the user volunteers them — contacts are added ""deliberately later on the People page, never auto-imported.\n\n"@@ -59,18 +67,26 @@ ONBOARDING_INTERVIEWER_SYSTEM = ("- EXACTLY ONE person has is_principal=true: the user themselves. If you ""are unsure which person that is, ask before drafting.\n""- Person full_name values must be unique.\n"- "- Every department's head_person_name MUST exactly match a person's "- "full_name; leave it empty when no leader is named.\n"- "- Department titles must be distinct. Model the functions this business "- "actually has — a 12-person startup does not have eight departments.\n"+ "- goals: what the user wants to achieve this year, one per item, in "+ "their words. owner_name is empty (the user) unless they said someone "+ "else owns it, and then it MUST exactly match that person's full_name. "+ "Also list the goals in profile.strategic_priorities.current_year.\n"+ "- departments: ONLY when the user's own words or files already list "+ "people by department (e.g. 'Dan runs engineering, Sara runs sales'). "+ "Otherwise leave departments empty — never invent them, and never make "+ "one per person. When given, every head_person_name MUST exactly match a "+ "person's full_name (or be empty), titles must be distinct, and a goal "+ "that clearly belongs to one names it in department_title.\n""- Use authority_level=propose_only unless the user explicitly says a ""function should act on its own.\n""- vendors and tickers are external dependencies the user named (a cloud ""provider, a key supplier, a public competitor or customer). Listing one ""authorizes the Executive to watch its public status and news without ""asking again, so include only what the user actually depends on.\n"- "- confidence_notes: one short line per field you could not determine, "- "saying what is missing. This is what the user sees flagged for follow-up.\n"+ "- confidence_notes: things the user should check, one plain sentence "+ "each, such as a guess you made or two sources that disagree. Write for "+ "a person: never use field names, and never list what the user chose "+ "not to share.\n""- summary: one short paragraph reading the company back to the user in ""plain language, so they can spot a misunderstanding at a glance.\n""- Numbers are numbers, not strings."diff --git a/packages/core/openexecutive/api/models.py b/packages/core/openexecutive/api/models.pyindex 99349cfc..917e9ce3 100644--- a/packages/core/openexecutive/api/models.py+++ b/packages/core/openexecutive/api/models.py@@ -517,6 +517,19 @@ class OnboardDepartmentDraft(BaseModel):authority_level: str = "propose_only"+class OnboardGoalDraft(BaseModel):+ """One of this year's goals: saved with an owner (a drafted person's+ name, empty = the owner of the workspace) and, when departments were+ drafted, optionally inside one."""++ model_config = ConfigDict(extra="ignore")++ title: str+ target: str = ""+ owner_name: str = ""+ department_title: str = ""++class OnboardTranscriptTurn(BaseModel):role: strtext: str@@ -530,9 +543,12 @@ class OnboardTurnResponse(BaseModel):max_questions: intquestion: str | None = Nonequestion_hint: str | None = None+ # Suggested answers to the question, shown as buttons to tap.+ question_options: list[str] = Field(default_factory=list)draft: CompanyProfileResponse | None = Nonedraft_people: list[OnboardPersonDraft] = Field(default_factory=list)draft_departments: list[OnboardDepartmentDraft] = Field(default_factory=list)+ draft_goals: list[OnboardGoalDraft] = Field(default_factory=list)confidence_notes: list[str] = Field(default_factory=list)summary: str | None = None@@ -549,6 +565,7 @@ class OnboardCommitRequest(BaseModel):profile: CompanyProfileUpdateRequestpeople: list[OnboardPersonDraft] = Field(default_factory=list)departments: list[OnboardDepartmentDraft] = Field(default_factory=list)+ goals: list[OnboardGoalDraft] = Field(default_factory=list)# The principal's sign-in email, confirmed by the user on the review# screen (pre-filled from their own login). The one contact detail setup# saves: without it the signed-in owner matches no Person, so their chatdiff --git a/packages/core/openexecutive/api/routes/onboarding.py b/packages/core/openexecutive/api/routes/onboarding.pyindex 4c1963ac..9730d18b 100644--- a/packages/core/openexecutive/api/routes/onboarding.py+++ b/packages/core/openexecutive/api/routes/onboarding.py@@ -24,6 +24,7 @@ from openexecutive.api.models import (OnboardAnswerRequest,OnboardCommitRequest,OnboardDepartmentDraft,+ OnboardGoalDraft,OnboardMessageRequest,OnboardPersonDraft,OnboardReviseRequest,@@ -294,6 +295,8 @@ class InterviewSession:# Hint for the LATEST question only. Lives on the session (not a side# dict) so it is swept with everything else and cannot leak.last_hint: str = ""+ # Suggested answers for the LATEST question only, like last_hint.+ last_options: list[str] = field(default_factory=list)saved: bool = Falselast_touched: float = field(default_factory=time.monotonic)# Wall-clock time of the stored copy this one matches (interview_store).@@ -386,6 +389,7 @@ def _turn_response(session_id: str, session: InterviewSession) -> OnboardTurnRes)for dep in d.departments],+ draft_goals=[OnboardGoalDraft(**g.model_dump()) for g in d.goals],confidence_notes=list(d.confidence_notes),summary=d.summary,)@@ -401,6 +405,7 @@ def _turn_response(session_id: str, session: InterviewSession) -> OnboardTurnResmax_questions=MAX_QUESTIONS,question=question or OPENING_PROMPT,question_hint=session.last_hint or None,+ question_options=list(session.last_options),)@@ -435,6 +440,7 @@ async def _advance_interview(if isinstance(result, CompanyDraft):session.draft = resultsession.last_hint = ""+ session.last_options = []# Record the draft as an assistant turn. Without it the transcript ends# on a user turn, and the next message would make two user turns in a# row — which the Messages API rejects, bricking the session.@@ -452,6 +458,7 @@ async def _advance_interview(session.transcript.append(Turn(role="assistant", text=result.question))session.questions_asked += 1session.last_hint = result.hint+ session.last_options = list(result.options)_persist_interview(session_id, session)return _turn_response(session_id, session)@@ -638,12 +645,14 @@ async def interview_revise(body: OnboardReviseRequest, request: Request) -> Onboquestions_asked=sum(1 for t in kept if t.role == "assistant"),draft=None,last_hint="",+ last_options=[],)response = await _advance_interview(body.session_id, trial)session.transcript = trial.transcriptsession.questions_asked = trial.questions_askedsession.draft = trial.draftsession.last_hint = trial.last_hint+ session.last_options = trial.last_optionssession.stored_at = trial.stored_atif session.discarded:# Start over landed while this turn was running: don't keep it.@@ -708,9 +717,10 @@ async def commit_interview(body: OnboardCommitRequest, request: Request) -> Compowner_change_blocked,owner_email_blocked,reconcile_onboarding_departments,+ save_onboarding_goals,save_onboarding_people,)- from openexecutive.onboarding.interview import DepartmentDraft, PersonDraft+ from openexecutive.onboarding.interview import DepartmentDraft, GoalDraft, PersonDraftsession = _get_interview(body.session_id, request)if session.saved:@@ -736,6 +746,7 @@ async def commit_interview(body: OnboardCommitRequest, request: Request) -> Comptry:people = [PersonDraft(**p.model_dump()) for p in body.people]departments = [DepartmentDraft(**d.model_dump()) for d in body.departments]+ goals = [GoalDraft(**g.model_dump()) for g in body.goals]merged = CompanyProfile().model_copy(update=update_data)profile = CompanyProfile.model_validate(merged.model_dump())except Exception as exc:@@ -757,7 +768,7 @@ async def commit_interview(body: OnboardCommitRequest, request: Request) -> Comp# all — find_principal_person() then returns None and caller resolution,# alert routing and the scheduler's brief have nobody to target.errors = validate_draft(- CompanyDraft(profile=profile, people=people, departments=departments)+ CompanyDraft(profile=profile, people=people, departments=departments, goals=goals))if errors:# .safe, never .detail — the detailed rendering quotes the rejected@@ -814,7 +825,7 @@ async def commit_interview(body: OnboardCommitRequest, request: Request) -> Comp),)- profile = derive_org_structure(profile, people, departments)+ profile = derive_org_structure(profile, people, departments, goals)# 2. is_empty() keys off the name — a nameless profile is invisible to# every consumer, including the health check that gates onboarding.@@ -847,6 +858,7 @@ async def commit_interview(body: OnboardCommitRequest, request: Request) -> Compif owner_email and principal_id is not None:link_owner_email(principal_id, owner_email)reconcile_onboarding_departments(departments, person_ids)+ save_onboarding_goals(goals, person_ids, principal_id)# 7. Same post-onboarding research fire the wizard does, same dedup set.if body.session_id not in _onboarding_research_fired:diff --git a/packages/core/openexecutive/architecture/architecture-facts.yaml b/packages/core/openexecutive/architecture/architecture-facts.yamlindex 99e1ee0b..d05781ef 100644--- a/packages/core/openexecutive/architecture/architecture-facts.yaml+++ b/packages/core/openexecutive/architecture/architecture-facts.yaml@@ -704,9 +704,8 @@ providers:executive_voices: |The Executive's voice persona (personas/builtin/*.md, shadowed byvoice_personas rows) is substituted into block 0 at {VOICE_PERSONA}.- Three general voices are offered by the shared VoicePicker in setup- (/onboard, after the workspace choice and solo role step) and Settings ->- Executive, saving voice_persona_slug through PATCH /agents/executive:+ Three general voices are offered by the VoicePicker in Settings ->+ Executive (setup no longer asks; it ends on Home), saving voice_persona_slug through PATCH /agents/executive:Direct (slug ``default``, stored as no override), Supportive andAnalytical. Frontmatter ``description`` / ``sample`` feed the picker;``legacy: true`` marks the named-executive voices, hidden from the picker@@ -9716,9 +9715,17 @@ company_onboarding:module: openexecutive.onboarding (interview.py, commit.py, wizard.py) + api.routes.onboardingintent: |First-run company setup. The user describes their business in their own- words (optionally attaching a deck, one-pager or brief); an LLM asks- clarifying questions, then drafts a company profile, a leadership roster- and a department list that the user reviews, edits inline, and saves.+ words (optionally attaching a deck, one-pager or brief); an LLM asks at+ most four quick questions, each with two to four suggested answers to tap+ (Question.options, MAX_OPTIONS), never about departments or finances, then+ drafts a company profile, the people, this year's goals with owners+ (GoalDraft) and departments only when the user's words already listed+ people by department. The user reviews, edits inline, and saves.+ Personal or Team is never asked: understand.py reads it (Team when the+ text names colleagues at the user's company, Personal when only the user+ or outsiders are named, Personal when unclear) and the UI saves it before+ the interview; the review shows it on one line with a Change link. Save+ goes to Home; there is no done screen or voice step in setup.Deliberately shaped like the /demo and /clients/generate draft-then-reviewflows rather than like a form.api: |@@ -9752,7 +9759,8 @@ company_onboarding:tool partway through a conversation invalidates the cached tool prefix onevery subsequent turn (the same invariant orchestrator/form_tools.pydocuments). tool_choice flips to the forced emit tool exactly once, on the- turn where the question budget (8) or the transcript ceiling (120k chars)+ turn where the question budget (MAX_QUESTIONS, 4) or the transcript+ ceiling (MAX_TRANSCRIPT_CHARS)is reached, or when the user clicks "draft it now". That single call missesthe tools cache; one miss per setup session is the accepted price of abounded interview. Do NOT "fix" it by making the tools conditional.@@ -9766,7 +9774,10 @@ company_onboarding:Two stages, mirroring fixtures.generator: Pydantic model_validate, theninterview.validate_draft for referential integrity — exactly oneis_principal, unique person names, every department head_person_name- matching a person (or empty), distinct department titles. A failure of+ matching a person (or empty), distinct department titles, and every goal+ named, within MAX_GOALS / MAX_GOAL_CHARS, its owner_name a drafted person+ (or empty) and its department_title a drafted department (or empty).+ The commit re-runs the same check on the body. A failure ofeither feeds the errors back as a follow-up user turn and retries ONCEwith the emit tool forced.error_contract: |@@ -9811,8 +9822,14 @@ company_onboarding:(commit.check_owner_email), (2) reject an empty name,(3) profile.save_to_yaml (write-then-rename, O_EXCL, fsync), (4) mark thesession saved, (5) seed people, then link the owner email to the- principal (best-effort), (6) reconcile departments, (7) fire the- post-onboarding research workflow. A failure at 1-3 leaves the session+ principal (best-effort), (6) reconcile departments, then save the goals+ (commit.save_onboarding_goals: period "year", no department unless one+ drafted is named, owner the named person else explicitly the principal,+ skipping any worded like a goal already set for this year or+ ongoing, so a re-run adds none; their+ titles also become strategic_priorities.current_year in step 3's+ profile via derive_org_structure), (7) fire the post-onboarding research+ workflow. A failure at 1-3 leaves the sessionuntouched, so the user edits the draft and retries. This replaces thedeepcopy-snapshot rollback the wizard needs (process_answer mutates statebefore the build can fail). A second commit returns 409, preserving thediff --git a/packages/core/openexecutive/architecture/prebuilt/agents.json b/packages/core/openexecutive/architecture/prebuilt/agents.jsonindex a904992e..9b428f2b 100644--- a/packages/core/openexecutive/architecture/prebuilt/agents.json+++ b/packages/core/openexecutive/architecture/prebuilt/agents.json@@ -1,7 +1,7 @@{"section_id": "agents","title": "Agent Council",- "markdown": "The Executive (`openexecutive.orchestrator.executive.Executive`) is a single coherent persona that fans out to specialist sub-agents via Anthropic tool use. Each specialist subclasses `openexecutive.agents.base.BaseAgent`, carrying a `name`, `domain`, and `model`, and exposes `analyze(...)`, which wraps the question in user-turn context and calls Claude through the per-call provider. That context is per call and never cached: `<company_stage>` (the stage from the company profile — specialists never see the profile itself), in solo mode `<principal_role>` (what kind of principal the advice is for, in plain words, with their title and remit — `router.principal_role_context`), `<department_memory>`, `<past_decisions>`, `<failure_cases>` (the shipped case studies, and under them the workspace's own lessons that match the call, `memory/lessons.py`), `<relevant_knowledge>` and `<conversation_context>`, outermost first. The specialist's own system prompt is a static domain constant under one `cache_control` block, so it stays byte-identical across companies and profile edits. A chat turn takes the stage from its session's profile (the one the Executive reasons over, and the one an eval scenario injects); workflow steps and the MCP server's `consult_specialist` read it fresh from the profile on disk. The role tag follows the turn's mode: a chat turn pins the role on its session with the mode (`pin_turn_principal_role`: the session's `principal_role` override, else the workspace settings) and sends it in solo and nothing in team, so the org block, every specialist in the turn and any workflow the turn starts see the same role; callers with no chat turn read it fresh (`load_principal_role`: the current session's mode and role, else the workspace's), and a failed read sends no tag rather than failing the consult. The CFO, CSO and sales prompts key their advice on it: for a bootstrapped, self-funded or solo company they drop venture fundraising benchmarks and work from cash, and raise only if the founder wants to.\n\nRouting uses one `consult_specialist` tool. When to call it is spelled out in the persona's Domain Expertise section (\"Going deeper\") and in the tool's own description: any question that asks for judgment in a specialist's area (a decision, plan, analysis, negotiation, hire, contract term, pricing or fundraising move, board message, or a department's own targets and stance, since the specialist holds that department's memory) is consulted before the answer, several at once across areas; lookups in mail, calendar or documents, scheduling, status checks, routine drafting and questions about how the Executive works are answered without one. Chat eval scenarios hold this: `required_routing` names specialists the turn must consult and `no_routing: true` says it must consult none, and the runner fails a scenario that misses either (`evals.scenarios.routing_check`). Beside it the Executive carries two skill tools, `search_skills` and `load_skill` (skills are shown to users as **playbooks**, and every playbook belongs to a workflow: one that follows it, or its own Quick workflow). Chat finds and reads playbooks but never creates or changes one: work worth repeating is offered as a workflow through `draft_workflow`, with the method written into its steps, because workflows read playbooks at run time (approved scheduled ones included) and this loop also runs on inbound email and chat, so only a person can change what a workflow follows. A hidden built-in is invisible to search and load; each `search_skills` hit lists the `workflows` that follow it, Quick ones included, and the persona routes \"the finished deliverable\" to that workflow and quick or partial asks to the playbook inline; Tool calls made while proactive work is tagged (scheduler dispatch, reflection, research, alert review) feed the Attunement outcome ledger. Goals have three tools: `list_department_goals`, `update_department_goal` (progress the user reports) and `create_goal` (a new goal the principal states, filed under an area by slug or title and creating the area when none matches — see Org). `create_goal` runs only for the principal on a verified surface, and no unattended run is offered it: the tools in `schedule_tools.UNATTENDED_WITHHELD_TOOLS` are dropped from reflection's and research's toolkits by `unattended_toolkit`, and from the chat loop when its Session is marked `unattended` — the scheduler's PROACTIVE TRIGGER run, whose prompt quotes stored intent text; a call the model emits anyway gets an error tool_result. `record_decision_outcome` records how a past decision turned out — the weekly review lists decisions older than 30 days with no outcome as `[decision N]` — writing `decisions.outcome` (`episodic.update_decision`) and auditing the rationale; it has `create_goal`'s gate (the principal on a verified surface only; everyone else gets an audited refusal) and is in `UNATTENDED_WITHHELD_TOOLS` too. A missing or unknown id returns the decisions still waiting on an outcome, so the model can retry with the right one. Its schema is static; the solo persona has one line on it, and team mode learns it from the tool description, so the team persona is byte-identical. Attunement adds `list_open_loops` (what a person owes — everyone's for the principal only in a private conversation, otherwise the caller's own) and `close_open_loop`, which only closes a loop when the turn's resolved speaker is the principal or the loop's owner — an unrostered sender, a signed-in web user not on the roster, and unattended sessions are refused. `assign_open_loop` opens one on request (\"ask Ben to send me the Q3 numbers by Friday\") for the principal or a rostered teammate on a verified surface — never an email, a turn private to the principal or an unattended run, which are also withheld the tool — and only when `task` appears verbatim in what the speaker typed this turn, at most 5 per turn; that turn's extraction pass then opens no second loop for the same owner. The owner is told once on their own channel when someone else assigns it, and the result's `notified` says how that went, so the Executive reports it rather than messaging them as well. The roster write tools (`upsert_person`, `archive_person`, `set_department_head`) are the principal's alone, and only on a surface that verified the speaker — the web chat, the principal's own Slack or Discord, a channel whose adapter verified the sender itself (`Session.speaker_verified`, with the `caller_person_id` it resolved, never on a turn an email started), or a private Telegram chat when `TELEGRAM_WEBHOOK_SECRET` is set to a value Telegram can send (Telegram identifies the chat rather than the sender, and without that secret its webhook accepts anyone's update). A teammate, an email (even from the principal's address), Google Chat, the CLI, the MCP server and unattended runs get `status: refused` (audited, and no action chip), because a team member's People row is also the web sign-in and outbound-email allow-list. `upsert_person` takes `kind` (`team` / `contact`; a new person defaults to `contact` in solo mode, an update keeps the current kind) and `email_aliases` (other addresses the person writes from — the full list, replacing the current one; they match mail and may be emailed but never sign in, and one already on someone else is refused), and after every write any pending roster request whose sender now matches someone is closed and its held messages replayed. `resolve_roster_request` answers a **roster request** — someone not on the People list who wrote in and is waiting (see Org): `approve` (a new person; `kind` defaults to team for a company-domain sender, else contact, when the principal did not say), `link` (someone already listed, by `person_id`; `replace_channel_id` only when the principal said so) or `decline`. It has the roster tools' gate (the principal on a verified surface; audited refusal otherwise), is in `UNATTENDED_WITHHELD_TOOLS`, and answers only an id in `Session.trusted_roster_request_ids` — the ids the server printed that turn in a `<roster_requests>` block (`briefing.context.format_roster_requests_for_prompt`, up to 10 pending requests), which `render_and_trust` appends to the `<briefing>` digest on the principal's own verified turn only and records on the session in the same step, so an id the model invents or reads out of a message adds nobody. Each line is server-derived — id, channel, address or account id, the sender's name sanitised and marked unverified, a company-domain hint, the message count — and never what they wrote. The block rides in the **user turn**, like the digest, so the cached system blocks are unchanged; the tool definition is static, and the new tool only extends the sorted tool list. The web chat route resets the trusted set every turn, and a resolved id is removed from it. and `list_people` returns both kinds with `kind` — contacts only on the principal's own verified turn (they are private to the principal; anyone else gets the team as if no contact existed), each with what is remembered about them from the principal's own mail and meetings (`what_you_deal_with`, `topics`, `recent_activity`, from `people.contact_notes`) so the Executive can suggest who to ask, and their contact details (`details`: phones, links, address), both on that turn only — while `lookup_person` searches the team only. `ask_about_person` on one of the principal's contacts, on the principal's own turn, returns that contact note instead of peer memory (contacts never chat, so peer memory has nothing on them); otherwise it reads a person's peer memory fully only on their own verified turn; anyone else's answers only to a verified principal or teammate, and only when that person turned on \"Share my work style with the team\", framed to how they work and never their view of a colleague, and otherwise returns the same \"hasn't chosen to share\" result whether or not anything is known. A teammate's own mail with the Executive is theirs the same way: every mailbox call the model makes through the MCP gateway (Gmail or Outlook, by `call_tool`, from a script, or a workflow step run from chat) passes through `orchestrator.teammate_mail`. A search excludes every other teammate's address server-side, except on mail the reader is also on, and is filtered again entry by entry; any other call that names a message or thread (read, attachment, reply, forward, label) must use an id a filtered search returned to the same reader in this conversation (`Session.mail_ids_seen`) or the inbound email the turn answers; reads are filtered message by message, so a message naming another teammate is left out unless the reader is in its From / To / Cc; a reply that would quote such a thread, or an Outlook forward, reply or attachment read of such a message, is refused; and a conversation in which a teammate other than the principal was shown their own mail becomes theirs alone, as Act as me's do. Outside mail and the reader's own are untouched; an email turn reads as its sender only when Gmail authenticated it, and any other unverified reader sees no teammate's mail, a background run the model drives reads as the principal and a workflow step outside any session as nobody, an Outlook inbox rule that forwards or names another teammate is refused, and the poller and the brief, which read the mailbox themselves, are unaffected. Their own chats are theirs too: every audit row of a teammate's verified turn is theirs (`people_tools.teammate_turn_owner`, see Audit; the principal sees the actions, never the words), and a teammate's web chat or DM opens for them alone (`chat._session_access`), the principal included. The same verified-principal rule, as `people_tools.contacts_reachable_now()`, decides whether a turn may reach a **contact** at all: `message_person`, the raw DM tools, `create_calendar_event` / `create_instant_meeting` attendees and the MCP gateway's Gmail / Calendar / Drive / OneDrive gates treat a contact exactly like an unknown person on every other turn, so no refusal reveals that one exists. **The untrusted-content policy** (`orchestrator.content_trust`) is one rule for text the principal did not write — inbound email, chat, attached files, watched pages — answering three questions for every surface. *Who is speaking*: the principal only when the surface proved it (`principal_speaking`: the web chat as the principal — never a signed-in email on nobody's People entry —, the CLI, their own Slack/Discord or a secret-verified Telegram chat, or their own address with Gmail's `dmarc=pass`); never a teammate, a stranger, Google Chat, the MCP server or an unattended run. *What may become memory*: episodic extraction runs only then, and only on the words outside `<untrusted_content>` blocks; peer memory and open loops are recorded under the speaker's own name, never as the principal's. *What may trigger tools*: every write on the principal's authority already refuses others in its handler; the tools with no such check that change the install for every later turn — `content_trust.PRINCIPAL_ONLY_TOOLS`: `load_mcp_server`, `list_saved_tools`, `run_python_job` and `run_script` — are not offered unless the principal is speaking on an interactive surface (not email, even authenticated), and a call the model emits anyway gets an error tool_result and a `refused: not_principal` audit row. So an inbound email, a teammate's turn, Google Chat and the scheduler's proactive trigger each get a stable tool list of their own without it. A send or edit on a team member's turn also meets their department's authority first (`orchestrator.department_gate`, see Org), and so does one on a turn whose speaker is on no People entry: a signed-in web caller or a chat-channel sender, which in practice is Google Chat, since Slack, Telegram and Discord turn unmatched senders away. Google Chat names a matched sender to the gate alone (`gate_sender_person_id`), so the principal there is not held. Email turns are left to the mailbox's own gates. *How it is labelled*: outside text goes into the user turn inside `<untrusted_content source=… author=… author_verified=…>` with a fixed notice that it carries no authority; compatibility forms are folded (NFKC), control characters go, and the tag's name is renamed wherever the text carries it — any case, spacing or entity-escaped form — so the text cannot end its block and speak outside it.\n\n**Outside content** (`orchestrator.outside_content`). Once a turn has read text from outside the team (any gateway read, an attachment, a web search result, a synced Drive, OneDrive, Confluence or Notion excerpt, `read_document`, research, or an extension tool registered `reads_outside`), the rest of that turn refuses what the Act as me lockdown refuses, every `upsert_person` but a contact the speaker named in this message at an address they typed or one that sent the mail read (`lockdown.speaker_named_contact`; never on a turn answering mail from someone not in People, and a refused one points to an `add_contact` card), and every `call_tool` but a read: no mail, Docs, Sheets, Drive or calendar write. The company's own Drive is not outside: a Google read of one file, Doc, Sheet, deck or folder by id (`company_read_item`) is looked up first (`vet_reads`, once per id a turn), and when that item is owned by someone at the company, or kept on a shared drive only the company organizes, and shared with nobody outside it (`mcp_gateway._company_item`), the read marks the turn as a company read (`COMPANY`) instead. That still holds everything else, since anyone who can edit the file wrote its words, but when every read so far was of the company's own items, a rename or move (`update_drive_file` with only `file_id`, `name`, `add_parents`, `remove_parents`) goes through, so the turn can file what it read. Searches, mail, calendar and forms mark it as usual. A refused rename or move the person asked for can go on a `drive_files` approval card instead. A round that reads also holds that round's other such calls; the read itself runs. Scripts are checked call by call. The offered tools never change, so the cached prefix stays the same, and the hold ends with the turn. Mail from someone not in People (a contact or a stranger) counts as outside content from the start of its turn (`Session.sender_off_team`), so the whole turn is held, with no direct messages or watchlist removals either; it can still raise an alert, private to the principal, which is how that mail reaches them. Someone added to People in a turn (or given a new address or chat account) is not emailed, messaged, invited or shared with until the next one. What a turn held is remembered for the conversation's next turn (`note_held`, in process, for an hour). It is kept per conversation and speaker. When that speaker's next message is a go (a short yes such as \"go\" or \"yes, do it\", or one opening with go, go ahead, do it or proceed as its own clause: `GO_RE`), the first round that reads has its reads refused once (`defers_read`, with an error that names no tool and asks for nothing new), so the steps the person just approved run before a new read holds them again; from the next round reads run as usual. It runs only what the model calls in that turn, under every usual check. A shared chat, and a turn answering mail from someone not in People, remember nothing. In the web chat the go is a tap: while it waits, the reply carries **Approve** and **Not now** buttons naming the held steps in plain words (`pending_held`, `step_words`: only names it knows; any other gateway tool shows its app's generic label, since the model's tool name can be steered by what it read), and Approve sends \"Go ahead\" as the person's own next message. The refusal on a web turn (`approve_hint`, set from `Session.from_web_chat`) tells the model to point them at Approve rather than ask them to say go; other surfaces still ask for a typed go.\n\nHow much a read holds is the owner's setting, **When I read outside content** (`orchestrator.outside_mode`, one row in `outside_content_mode`, changed only by the principal through `GET`/`PUT /executive/outside-content` on Settings → Your Executive), fixed per turn in `begin_turn` (`TurnOutside.mode`). **Ask every time** (`ask`) is the hold above. **Only for outsiders** (`outsiders`, the default) skips what came from the team: a Gmail search that shows only ids and links, a Gmail message read where every message's sender is a teammate (not a contact) and Gmail authenticated it (`fact_confirmation.sender_authenticated` on a raw read of each id, at most 10) and no other address appears, and a calendar read naming only the team where the Executive itself organized every event (a teammate's invite counts as outside, since its organizer is whatever the mail that brought it says) (Google's `get_events` only when `detailed` or for one event, since the plain listing doesn't say who made each event; Microsoft 365's calendar reads) (`mark_read` / `from_the_team`, checked on the result; an `@` the address pattern can't read counts as a stranger). Attachments and synced folders (a file carries whoever wrote it, not whoever sent it), web pages, fetched documents, vendor apps, Outlook mail and every other gateway read still hold. **Don't ask** (`trust`) applies only on the principal's own turn on an interactive surface that isn't email or a shared chat (`content_trust.principal_only_withheld` empty): no read marks the turn, and reading their own mail holds nothing either (the Act as me lockdown and `app_control` are skipped via `trusted_turn`); anyone else's turn gets `outsiders`. Whatever the setting, a shared chat, another AI app's turn and mail from someone not in People are `ask`, sends still go only to people on People, and someone added this turn is reached no earlier than the next one. An unreadable setting is `ask`.\n\nA turn about the principal's private mail (`Session.private_to_principal`) reaches the principal only. It is neither offered nor allowed to run `schedule_tools.PRIVATE_TURN_WITHHELD_TOOLS` — everything that reaches someone else, publishes where the team reads, or starts work outside the turn: the broadcasts, meeting bookings and cancellations, `run_workflow` and research, follow-ups and workflow nudges, watches, artifacts, goal updates, saved workflows, skill drafts, and `load_mcp_server` (it connects to any URL, which can carry the turn's content). Through `search_tools` and `call_tool` it reaches only `schedule_tools.PRIVATE_TURN_MCP_TOOLS`: Google Workspace reads (mail, calendar, and Drive, Docs and Sheets search and file reads, with their Microsoft 365 twins; OneDrive's `download-bytes` only for a file's content) and the Gmail send and draft the gateway narrows to the principal. A search shows it nothing else, and a call to anything else (Google Chat, a Docs or Drive write, `manage_event`, another server) is refused (`schedule_tools.private_turn_withholds`). A call the model emits anyway, MCP tools included, gets an error tool_result and writes an audit row, and every audit row the turn writes is the principal's alone.\n\nA turn another AI app asks over the MCP server (`Session.from_mcp_client`, set by `ask_executive`) reads and advises but never acts, unless the request may act (below). It is neither offered nor allowed to run `schedule_tools.MCP_CLIENT_WITHHELD_TOOLS`: the private-turn and unattended sets, plus every direct message (`message_person`, the Slack, Discord and Telegram DMs), alert acks and new alerts, brief edits, closing open loops, roster and department-head writes, watchlist changes, workflow drafts and form proposals. It gets no Act as me, `recall_history` or extension tools, and of the gateway tools only `schedule_tools.MCP_CLIENT_MCP_TOOLS`: the private-turn reads without the Gmail and Outlook send and draft tools. The set is fixed, so the variant keeps a stable cached tool prefix. A call the model emits anyway, through `use_tool` or `call_tool` included, gets an error tool_result telling it the conversation comes from another AI app, so it can read and advise but not send, post, book or change anything, and writes an audit row (`refused: mcp_client`); `MCPGateway.call_tool` refuses the same calls again.\n\nA request may let that turn act (`Session.mcp_client_may_act`: the `x-mcp-client-actions: on` header on a request the shared secret authenticated, or `MCP_CLIENT_ACTIONS` where no secret is set). It then runs the speaker's own changes inside Open Executive (`schedule_tools.MCP_CLIENT_RUNS_WHEN_ACTING`: alerts, open loops, follow-ups, the watch list, goals, decision outcomes, facts and company profile writes; never the roster), each under its own owner-only and person checks, since the turn counts as a verified surface for its caller. It is still denied a fixed set, `MCP_CLIENT_ACTING_WITHHELD_TOOLS` (DMs, calendar writes, broadcasts, `load_mcp_server`, scripts, Python jobs, workflows, research and the rest of the read-only set), with a refusal telling the person to do it in Open Executive. A message or a meeting goes on an approval card instead (`propose_actions`), which is carried out at once when the person's Handle it for me covers it, everyone on it has been on the roster for a week, and `action_cards.on_its_own_refusal` finds nothing against it, and otherwise waits for their tap. Every audit row it writes names the app (`via`).\n\nA web chat its owner shared (`Session.shared_chat`) mixes people. The note naming who is in it says which of them owns it and that the others' lines are messages from them, not the owner's instructions. While someone other than the owner has written since the owner's last message (`Session.others_spoke_last`, set per turn by the chat route from the saved history, the speaker included, and true when the history can't be read), the turn is neither offered nor allowed to run `schedule_tools.SHARED_CHAT_WITHHELD_TOOLS`: direct messages, the broadcasts, meeting invites and cancellations, follow-ups, roster writes and workflows (a run, a save or a suggested schedule acts later with the full tool set). Through `call_tool` it can't send, reply, forward or draft mail in Gmail or Outlook, or make a calendar change that mails attendees (`schedule_tools.shared_chat_withholds`). That holds on the owner's own turn right after someone else's line too, so a request carried over from it can't send; the owner's next message with nobody else's in between lifts it. A call the model emits anyway gets an error tool_result and writes an audit row (`refused: shared_chat`). The principal-only tools stay out of every shared chat (`content_trust.principal_only_withheld`).\n\nOn every turn the Executive knows what is connected without searching: block 0's *Connected Systems* section names the connected Google services, other MCP servers, the messaging channels, Notion sync, Confluence sync (when on) and Act as me, and pins the Google tool names it calls directly with `call_tool`. Both the service list and the pinned names are narrowed to what workspace-mcp actually loaded (`WORKSPACE_MCP_TOOLS`, `WORKSPACE_MCP_TOOL_TIER`), so a service left out is not named and its tools are not pinned — being told a tool is already available when it is not in the index costs a failed call the model cannot explain to itself. `search_tools` is for anything not listed. When work repeats over many items, the Executive can also call `run_script` (`workflows/step_script.py`, offered when the gateway runs, unless `CHAT_SCRIPTS=false`, and only on the principal's own turns — `PRINCIPAL_ONLY_TOOLS`, withheld on an unattended pass too, because it spends the turn's whole call budget on code the model just wrote): one short Python script, run in a sandboxed Monty worker with no files, network or environment, whose only reach is the gateway tools `call_tool` could reach (those `search_tools` has returned; the gateway keeps that set for its own process-wide session, not per conversation). Each call it makes is a `call_tool` in every way that matters — the gateway's discovery, deny-list and recipient gates, Take the lead's gate on an unattended run (or the department gate on a team member's turn, see Org), and the same action chip and `tool_invocation` row (marked `via: run_script`, `ok: false` when it raised), as each call happens; the script itself gets a `kind: script` row with its source and result. Its wall clock is half the turn's deadline, so a long script ends with the list of calls it made rather than being cut off with the turn, and a stopped turn closes it and its worker. A turn private to the principal is not offered it, and the Act as me lockdown withholds it whole. Besides gateway tools, a chat script may call a fixed list of the Executive's own tools (`step_script.CHAT_OWN_TOOLS`: reads, alerts, open loops, goals, the watch list, follow-ups, `message_person`, `upsert_person` — nothing that broadcasts, books calendars, edits memory or skills, or starts workflows or research), each only when the turn is offered it and through the turn's own handler (Take the lead's gate included), with the same chip and a `kind: skill`, `via: run_script` audit row. A direct call needs a `tool_use` each, so a turn's scripts share a budget instead: `CHAT_SCRIPT_MAX_CALLS` (default 200) calls in all, and per-turn caps on own tools that could otherwise act in bulk (`CHAT_OWN_TOOL_CAPS`: `message_person` 10 — messaging the whole roster would be a broadcast — `upsert_person` 25, `create_alert` 50); a refused call tells the model to say what is left. A call stopped mid-flight by the script's clock or the turn still gets a `cancelled` audit row, since it may have run. Because every round re-reads the conversation, the loop also nudges at the moment of choice: after a round whose results include a list of five or more items (`step_script.lists_many`), with `run_script` on offer and not used that round, it adds the constant `FANOUT_HINT` as a text block in that round's user message (never a cached block), once a turn, pointing the per-item work at one built tool. For library work on files (PDFs, spreadsheets, charts, images) the chat also has `run_python_job` (see Workflows → Python jobs): Python with a fixed library set in a Deno-isolated WebAssembly sandbox, fed Knowledge files by name, the message's own attachments as sent, or text it already holds, and able to keep a job that worked as a Custom tool, with result files linked for the principal to download; it is offered only on the principal's own turns and only where the sandbox is installed (not offered otherwise, so the tool prefix stays fixed per process). Agent Activity shows each run as one \"Built tool\" card (debug event `script_run`: ok, the calls it made, how many, how long, and the kept or reused tool's name). The persona's tool-building section is present only when chat scripts can actually run (the setting and Monty installed). `TOOL_BUILDING_ADDENDUM` in the cached persona tells the Executive when to build a tool — a job that looks something up and then checks or acts on each result, paged results, more than about ten items, or an answer that needs three or more reads (find, open, read; the tool returns only what answers the question, not whole documents, and only what is about to be edited or quoted at length is read in full), while a handful of items already named go as direct calls in one turn — and never to say it has none, and to talk about it as building a tool or doing it in one go. On the principal's own verified, interactive turns, a script that works can be saved as a named tool (`save_as`; elsewhere the script runs but is not saved) and run again with `run_script(tool=…, inputs=…)`; `list_saved_tools` (a constant definition, in `PRINCIPAL_ONLY_TOOLS` with `load_mcp_server`, so offered only while the principal is speaking on a verified, interactive surface; `run_script(tool=…)` is refused on any other turn) lists them, and each call a saved tool makes is a gateway call like any other (see Workflows → Saved tools). The tool's definition is a constant, so the cached tool prefix does not move. Multiple parallel `tool_use` blocks in a single assistant turn fan out concurrently through `router.route_parallel`, each running its own domain-filtered ChromaDB query (3 built-in and 2 company chunks by default, see the RAG section) before the specialist sees the question. The fan-out is failure-isolated per call: a specialist that raises or returns no text becomes an `UNAVAILABLE` tool_result (`router.specialist_unavailable_result`) instead of failing the whole turn, and its area — \"finance\", \"legal\", never the specialist's name — is reported to the user under the web reply, or by the Executive in its reply on other channels, unless another call to it answered that turn. A failed specialist does not count as consulted, so it takes no committee seat. Cancellation (Stop) still propagates. The Executive then synthesizes the returned analyses into one reply.\n\n`remember_fact`, `forget_fact` and `update_company_profile` (`orchestrator/fact_tools.py`; the last two sit in the `memory` group of `open_tools`, since they are rarely needed) make a correction stick: the first two keep or retire a standing fact in `memory.facts`, and the third changes one field of the company profile. They share `record_decision_outcome`'s gate (the principal on a verified surface, an audited refusal for everyone else) and are in `UNATTENDED_WITHHELD_TOOLS`. `remember_fact` alone also admits a rostered teammate on a verified surface (web signed in, their own Slack or Discord): the fact is stored attributed (`recorded_by_role=\"teammate\"`, rendered `(per <name>)`), and is held as a `proposed` row (`awaiting_approval`) until the principal approves it, unless the principal marked that teammate trusted (\"needs my approval\" off; on by default) — and even then when it would replace the principal's fact. All three are also in `PRIVATE_TURN_WITHHELD_TOOLS`, since what they write or retire is read on everyone's turns. On the principal's own authenticated email (primary address, with Gmail's own `Authentication-Results` reporting `dmarc=pass`, not forwarded) the three tools check the request as usual, then hold it as `awaiting_confirmation`. It applies only when the principal replies to a one-time token emailed to their own address (`integrations.fact_confirmation`); the split into a check step and an `apply_*` step lets chat and the confirmed reply share one apply path. Each write also needs a `source_quote` found in the speaker's own words this turn. The active facts reach every specialist as a `<standing_facts>` user-turn tag: `route_to_specialist` reads them (None → the store, \"\" → none) for chat consults and workflow steps alike, and `BaseAgent.analyze` places the tag next to `<past_decisions>`, never in the cached specialist system prompt.\n\nThe ten entries in `SPECIALIST_REGISTRY` (`openexecutive.orchestrator.router`):\n\n| Key | Agent | Domain |\n| --- | --- | --- |\n| `cso` | `StrategyAgent` | strategy |\n| `cfo` | `FinanceAgent` | finance |\n| `chro` | `HRAgent` | hr |\n| `gc` | `LegalAgent` | legal |\n| `coo` | `OperationsAgent` | operations |\n| `cmo` | `MarketingAgent` | marketing |\n| `cpo` | `ProductAgent` | product |\n| `sales` | `SalesAgent` | sales |\n| `board_comms` | `BoardCommsAgent` | board |\n| `triage` | `TriageAgent` | triage |\n\nModels: specialists default to `claude-sonnet-5-5`; the deep-reasoning tier (CSO, CFO, GC, Board) runs on `DEEP_REASONING_MODEL`, `claude-sonnet-5-5` by default, with adaptive thinking; routing/fast classification uses `claude-haiku-5-5`. Requests are fitted to the model's generation on the way out (`feature_gate.fit_claude_generation`): 5-generation Claude models never receive `temperature` / `top_p` / `top_k`, and Haiku 5.5, which thinks unless told otherwise, gets thinking turned off when the caller asked for none. The Council's **Deep reasoning** toggle is honoured on Claude Opus/Sonnet (direct or via OpenRouter) and on any OpenRouter model whose live catalog entry supports `reasoning` (today all OpenAI, Gemini, DeepSeek and Grok entries; not Llama) — the request translator turns Anthropic adaptive thinking + effort into OpenRouter's `reasoning` parameter. It is a silent no-op on models that can't reason, and disabled for Haiku, which rejects adaptive thinking. A PDF `document` block is translated too: it becomes an OpenAI `file` part, and OpenRouter requests carrying one add the `file-parser` plugin (`native` for models whose catalog entry lists `file` input, else `PDF_OPENROUTER_ENGINE`, default `mistral-ocr`). `FeatureSpec.supports_pdf_input` gates it per model — off for a local server unless `LOCAL_PDF_INPUT` — and a gated-off PDF becomes a text note rather than vanishing. The Council picks a model in two steps: a **Provider** dropdown (Anthropic, OpenAI, Google, Meta, DeepSeek, xAI, Local, …), then a **Model** dropdown limited to that provider, with readable names (\"Claude Opus 5.5\") and the raw id and serving route shown underneath. The Anthropic group combines the direct Claude models (`claude-fable-5-1`, `claude-opus-5-5`, `claude-opus-5`, `claude-sonnet-5-5`, `claude-sonnet-5`, `claude-haiku-5-5`, `claude-haiku-4-5`) with any Claude generation that exists only on OpenRouter, and never lists a model twice. With `OPENROUTER_ENABLED`, the non-Anthropic entries come from OpenRouter's live `/models` catalog (fetched at startup, refreshed in the background, filtered to tool-capable paid text models from an allowlisted vendor set), falling back to a built-in snapshot if the fetch fails. Every Council agent also takes **additional instructions** (`agent_overrides.instructions`): text appended after its prompt by `agents.overrides.append_instructions` inside an `<additional_instructions>` block. `BaseAgent.base_system_prompt()` is the built-in prompt or its full replacement (what the Council's editor shows); `effective_system_prompt()` is that plus the instructions (what every call sends), and the Executive gets the same text through `build_system_blocks(persona_instructions=…)`. Unlike a replaced prompt, instructions leave the agent on its built-in prompt, so prompt improvements keep reaching it. A **Quality** choice above the agents (`agents.presets`) sets every agent's model and deep reasoning at once: **Fast** puts every agent on a small model (`claude-haiku-5-5`, else `claude-haiku-4-5`) with deep reasoning off, **Balanced** returns each agent to its built-in defaults, and **Thorough** puts the Executive and the domain specialists on the largest allowed model (`claude-opus-5-5`, else `claude-opus-5`) with deep reasoning on where the model supports it, leaving triage and the helpers below on their defaults. Presets change quality only; they never cap spend. Each preset resolves against the install's model allowlist at request time and is offered only when one of its models is allowed. Applying one writes ordinary per-agent overrides (model and deep reasoning only), so history, rollback and later per-agent edits keep working; an agent changed on its own afterwards shows as **Custom**. Each Council agent has a **visibility**: `core` for the Executive and the domain specialists (`BaseAgent.visibility = \"core\"`, which a new specialist must declare) and `internal` (the default) for triage and the helper agents below. Thorough moves only core agents. The Council opens in a **simple view**: the Quality choice, the voice picker, and only the core agents, each with its additional instructions and a model badge. **Show all agents** lists the internal ones too, and **Advanced** opens the full page (the browser remembers the choice). Eight Council-only agents are model/prompt-overridable in the UI but NOT callable via `consult_specialist`, because they are deliberately kept out of `SPECIALIST_REGISTRY`: `quality_judge` (committee review), `utility_fast` (gates, titles, decision parsing), `research` (the `executive_research` fan-out), `fixture_generator` (invents demo companies), `engagement_intake` (models a real client from intake material), `onboarding_interviewer` (interviews a new user about their own company and drafts its profile, people and departments), `workflow_designer` (runs the conversational New workflow wizard on `/jobs/new`: searches the available tools, asks clarifying questions, then drafts a validated dynamic-workflow definition for the user to review), and `workflow_actor` (runs a workflow's `action` steps: a bounded tool-use loop limited to the tools the user approved for that step). `fixture_generator`, `engagement_intake` and `onboarding_interviewer` share one shape — a forced tool call producing a whole company, validated for referential integrity, with a single repair retry — but opposite epistemics: the fixture author invents, the other two extract and leave unstated facts empty rather than guessing. When the workspace is in solo mode, `onboarding_interviewer`'s first user turn is prefixed with a short hint (`interview.solo_hint(role_kind)`, one static constant per role kind from the onboarding role step): the draft names only the principal, as principal, and no departments; an owner is asked about the offer, customers, pricing, cash and runway, an in-house executive about their organisation (drafted as the company — their employer), function, remit, manager, team size as context and what they are measured on, an independent about their practice, clients and pricing, and with no role given the role-neutral `SOLO_HINT` asks which they are. The hint rides in the user turn and the tool array is unchanged, so the cached system prompt and tool prefix are the same in both modes.\n\n**Voice.** The Executive answers in one voice persona, substituted into block 0 at `{VOICE_PERSONA}` (`personas/loader.get_active_body`) from the Executive override's `voice_persona_slug`. Three general voices ship in `personas/builtin/`: **Direct** (`default`, the slug used when none is set: answer first, brief and decisive), **Supportive** (a warm coach who explains the why and checks how the plan sits) and **Analytical** (numbers first, options and trade-offs, then one recommendation). They are offered by a shared picker in setup (`/onboard`, after the workspace choice) and in Settings → Executive, which saves through `PATCH /agents/executive`; a built-in voice's frontmatter carries the picker's one-line `description` and `sample` reply. The named-executive voices are marked `legacy: true`: the picker hides them, but they still load, so an install that picked one keeps it, and the Agent Council lists them under \"Legacy voices\". Voice files hold disposition only; the response-length ladder lives in `prompts/executive_persona.py`, and a test keeps numeric length rules out of every voice file. An eval scenario may set `voice_persona` to answer in a voice; it rides on `Session.voice_persona_slug`, which wins over the override for that session only.\n\nOne more structured-output utility agent sits beside `triage`: `openexecutive.agents.alert_review.AlertReviewAgent` (routing model, one forced `emit_alert_reviews` tool call, never raises). The alert review job hands it a batch of open alerts with evidence and gets back, per alert, a verdict (relevant / changed / resolved / stale), a confidence, the evidence it relied on, a one-line note, why-now, an optional deadline, and a recommended move (none / route / nudge / escalate / draft / suggest_workflow / merge / close) with its parameters. Like `triage`, it never mutates state — `alerts.review.apply_verdict` executes the verdicts deterministically within authority, so the model's job is judgment, not action.\n\n**Goals without departments.** The goal tools (`orchestrator.department_tools`) no longer assume a department: `list_department_goals` with no `department_slug` lists every goal with its owner and department, `update_department_goal` names a goal by `goal_id` alone, and `create_goal` takes an optional `area` and `owner_person_id` (a team member). The org block renders goals with no department with their `goal_id`, and their action chips link to `/goals` (see Org).\n\n**Solo mode.** When the workspace (or the turn's session) is solo — one person using Open Executive for themselves — the Executive works for that one person, the principal, whatever their role: they may run their own business, lead a function inside a larger organisation, or work independently, and the persona reads which from the context instead of assuming it — the role lines under \"Your Principal\" in the org block, when the principal gave their role (workspace settings: kind, title, reports to, remit, measured on). Its persona swaps the team sections (\"You Are a Member of This Executive Team\", \"Choosing Who to Tell\", \"Departments & Org Coordination\") for \"You Work for One Person\", \"The People in Their World\", \"Who Hears From You\" and \"Goals and Areas\". Solo means only the principal uses Open Executive — not that they have no people: their manager, direct reports, peers, clients, investors and board are contacts. The Executive helps the principal work with them (a 1:1 or review prep, an update or budget case for their manager or board, a negotiation) but does not coordinate them, contacts them only when the principal asks and only if they are a contact, and never offers to loop the principal in. Every specialist stays relevant — HR for the principal's own reports, board and investor communications for reporting upward. Goals are grouped by *area* (the tools' `department_slug` is the area's slug). The chat loop withholds `send_department_message`, `send_company_broadcast` and `set_department_head` from its tool list, and a call to one of them anyway gets a JSON error tool_result instead of running. The mode is resolved once per turn and pinned on the session, so the persona, the org block, the tools and every tool handler agree — a switch made mid-turn applies from the next turn. Team mode is unchanged.\n\nEvery specialist *advises* — each returns domain analysis the Executive synthesizes into one voice. None of them act on the company's behalf; the side-effecting tools in the main loop (schedule, people, departments, broadcast, watchlist, workflows) belong to the Executive, not to a specialist.\n\n`recall_history` (Always in the loop, `orchestrator.history_tools`) reads the speaker's own notes, on the same terms as `ghostwrite_email` below: its own registry, offered only on a verified turn in a conversation private to them with their switch on; its `due` flag lists only dated notes, soonest first (see Memory). The ghostwriter below also gets the writer's notes from mail every recipient was on, as a consistency check (`<writer_noted>`).\n\nThe morning reflection, which has no question to look up and no specialist to consult, gets one more read-only tool of its own: `search_knowledge` (`orchestrator.knowledge_tools`), the same knowledge lookup a chat turn makes, for what the company's documents say about a customer, contract, price or policy before it acts.\n\nA fifth, `remind_me` (`orchestrator.reminder_tools`), sets a plain-text reminder sent to the speaker alone when it is due; it opens no mailbox, so it never locks a turn. A sixth, `propose_actions` (`orchestrator.action_card_tools`), does nothing itself: it leaves the speaker a card of up to five actions (a meeting, a message, a new contact) that happen only when they tap Approve, so it stays on after a turn read mail; it is where what an email itself asks for goes, since that is never done on the email's say-so. Four read-only tools beside it read the speaker's own mailbox on the same terms — `search_my_email`, `read_my_email`, `read_my_email_attachment`, `my_email_awaiting_reply` (`orchestrator.mail_read_tools`) — and `my_calendar` their own calendar while Read my calendar is on, capped per turn, other people's messages and attached files returned as untrusted blocks (see Act as me). Beside them, `save_my_email_attachment` copies one attached file, when the speaker asked in their own words, into a fixed folder of the Executive's own Drive or OneDrive and shares it with them alone, every call built in codeOne tool writes under someone else's name, and it is fenced in code rather than the prompt. `ghostwrite_email` (**Act as me**, `orchestrator.delegation_tools`) writes an email as the person the Executive is speaking with and saves it as a draft in their own mailbox, Gmail or Outlook (`delegation.gmail.gmail_for` picks the client by the person's credential); chat never sends it (the person sends it themselves, or taps Send on its card). It lives in its own registry, never `_ALL_SKILL_TOOLS`, so reflection, research and workflow toolkits can never carry it. The chat loop adds it only on a turn `delegation.settings.pin_turn_delegation` offered it to — the principal, or a team member once the owner lets them, with it on, verified, in a conversation private to them (the web chat, a Slack or Discord DM, a verified private Telegram chat), on an attended turn — and builds that turn's handler map from both registries, so on any other turn a call to it is an unknown tool. From the round that calls it until the turn ends, the dispatch guard refuses only what opens an outside address, runs a script or workflow, or posts to everyone (`delegation.lockdown`), without changing the offered list; messages and invites to the roster, follow-ups, facts in the person's own words and a contact they name still run, each fenced in its own path. A later turn in the conversation stays private to its owner but holds nothing: the person asking again, having read the reply, is their own request. A contact add the lockdown refuses becomes an approval card in the chat instead (`propose_actions`). Recipients are chosen in code, never by the model (a draft's new email may go to any address, named back to check when it is neither on the roster nor typed), and the text comes from the **ghostwriter** (`delegation.ghostwriter`), a tool-less call with a constant prompt plus the person's \"How I write\" profile. The ghostwriter is deliberately not a Council agent: the Council's edit routes have no principal check. See **Act as Me (Delegation)**. A second tool joins the same way: `recall_history` (`orchestrator.history_tools`, Always in the loop) reads the speaker's own notes of what they said, by email or in chat. It has its own registry and per-turn handler map too, and the loop adds it only when `history_tools.recall_person` finds the speaker verified, in a conversation private to them, on an interactive turn that isn't private to the principal, with their notes switch on. It doesn't need Act as me. After every chat turn the Executive also hands the speaker's own words to `history_chat.schedule_chat_notes`, which takes notes in the background when their switch is on and the surface verified them (see **Memory**). A meeting `create_calendar_event` proposes works the same way when the person chatting is the one who approves it: its chip carries the `decision_id` and the chat shows the meeting card (Approve, Edit, Dismiss) under the reply, while the briefing keeps it too. A tool result held for someone else's approval (`waiting_for_approval`, the department gate) gets a *Waiting for approval* chip, never a done one.\n\nThe Executive can also launch *any* workflow conversationally: `orchestrator.workflow_run_tools` adds `list_workflows` (the launchable catalog with each workflow's inputs) and `run_workflow` (runs a built-in or custom workflow to completion and returns its artifact, checkpointing approval gates as `awaiting_human`). `run_workflow` runs the weekly review, and in solo mode the morning brief, only for the principal on a verified surface (`Workflow.principal_only_modes`, `create_goal`'s rule); anyone else gets an audited refusal that does not name the rule. A run whose `result` event reports `private_to_principal` (the morning brief on the principal's own verified turn, when it read their private rows) still returns its artifact into that turn, but the stored run keeps `persistence.PRIVATE_RUN_ARTIFACT` instead of the text, since run history is readable by everyone signed in. This makes the `/jobs` catalog reachable from chat — see the Workflows section.\n\nDeliverables are first-class too. `orchestrator.artifact_tools` gives the Executive three tools: `draft_artifact` publishes a memo, web page (`html`), Word document (`docx`), spreadsheet (`xlsx`, from structured `sheets`) or a `link` to something it created in one of the principal's connected apps through the generic MCP gateway (`search_tools` → `call_tool`, any server — no vendor is named), and returns the artifact's id and link; `list_artifacts` and `get_artifact` let it find and reread its own past work (drafts and workflow outputs alike, through the same `artifact_records` read model the API uses). Every artifact belongs to someone: a draft to the person whose turn published it (`artifact_records.current_viewer`; with nobody speaking, the scheduler or the CLI, the principal), and a run someone started by hand to them, while scheduled runs are the team's. The tools, the recall, `message_person`'s link and mail attachments only ever reach the speaker's own (a link is refused when the recipient couldn't open it), not even the principal sees a teammate's, and a speaker who isn't on the People list publishes nothing. A workflow's own session inside someone's turn (the research synthesis) is pinned to that turn's viewer (`Session.documents_viewer`). Revising is `draft_artifact(supersedes=<id>)`, which archives the earlier version. When the principal asks for a deliverable the tool is always used; the \"quiet is the right default\" bar applies only to unprompted drafts. The format enum is a hard-coded sorted tuple so the cached tool schema stays byte-stable. Delivery: `message_person(person_id, text, artifact_id?)` appends the artifact's title and a `UI_BASE_URL/artifacts/<id>` link (through the same outbound guard as any DM; the link needs Open Executive access), and a Gmail send/draft can attach the file itself by `artifact_id` — see Integrations.\n\n**Extensions** (`orchestrator.extensions`). An install can add chat tools and kinds of document without changing this code. `OPENEXECUTIVE_EXTENSIONS` names Python modules. The first turn, or the first Documents read, imports each one and calls its `register()`. A module that fails is logged and skipped. `register_tool` takes an `ExtraTool`: a definition, an async handler, an optional `tool_groups` group (new or existing) and an optional chip. It refuses a name the loop already knows (`executive.builtin_tool_names`). An extension tool is never in `_ALL_SKILL_TOOLS`, so reflection, research and the synthesis never see it. The chat loop adds it per turn, like the Act as me tools, but never on an unattended run, a turn private to the principal or to a speaker who isn't on the People list (mail from a stranger), and only where its own `offered(session)` says. A tool registered `known_caller=False` is the one exception to the People-list rule: it is asked on a turn with no verified caller too (a surface of the extension's own that names its user without proving who is speaking), so `register_tool` requires it to have an `offered` that checks who is asking; it is still never offered on an unattended, private, AI-app or replay turn. That predicate depends only on the kind of turn, so each kind of turn keeps one cached tool list. Act as me's lockdown refuses it once the turn has read the owner's mail, because it is unclassified and the lockdown fails closed, unless it was registered with `after_mail_read=True`: a tool that only reads the install's own data and carries nothing the model wrote anywhere outside. Its chip goes through `action_chips.summarize_action`, which applies the same no-chip rules for an error or refusal and keeps a link to a path inside the app. A `waiting_for_approval` result is the tool's own chip's to describe, never core's generic \"Waiting for approval\": the department gate never holds an extension tool. A tool may also name the chat's activity line while it runs (`activity_label`) and keep its dispatch audit row private (`private_row`: its input or result quotes the speaker's own data). `register_collection` adds a kind of document: a drafted artifact tagged `collection:<name>` carries that collection on its record. The Documents page gives each collection its own filter. The collection's `on_change` hears about archive, restore and delete before they happen; if it raises, the change does not happen. `draft_artifact(supersedes=…)` refuses a collection's document. The alert pipeline strips `collection:` tags, so only the extension's own code files a document in its collection. A module that fails partway through `register()` leaves nothing registered. A tool registered with `workflow=True` is also offered to workflow action steps as `oe__<name>` (see Workflows); `read_only=True` marks one that only reads. A step runs with nobody in the turn, so `offered` isn't asked: `extensions.in_workflow_step(name)` is True for that tool, during its call only, and the handler decides what an unattended call may do.\n\nDocuments come in the other way through `read_document` (`orchestrator.document_tools`): it reads a PDF, Word, Excel, CSV, Markdown or text file by `path` — one another tool saved, such as a Gmail attachment (`get_gmail_attachment_content`) or a Drive file (`get_drive_file_download_url`) — or a company document by `filename`, with an optional PDF `pages` range. A scanned PDF is converted (`knowledge.pdf_reader`: local OCR by default; with `PDF_PROVIDER_READING=true`, the deployment's own model through its provider's PDF support, falling back to local OCR) rather than returned empty. Reads are confined to the tool download folders — the same resolve-then-contain check as `oe__read_file` — plus the company docs folder, and it is withheld on turns private to the principal (`PRIVATE_TURN_WITHHELD_TOOLS`): those come from a contact's email, whose own attachments the poller has already read into the turn, and must not steer the Executive into company documents.\n\nThe Executive also backs the **Ask OE** panel's form-fill tier via `orchestrator.form_tools.propose_form_values`: when a chat turn arrives with a `<page_context>` block describing a form on the user's screen, the Executive can propose values for it. The proposal is delivered to the UI as a `form_patch` SSE event and placed into the form as highlighted suggestions — the tool never saves or submits anything; the user reviews and saves manually. The tool is **always** in the tool list (tool definitions sit at the top of the cached prompt prefix, so a conditionally-present tool would invalidate the prompt cache whenever panel and non-panel turns alternate); its description restricts usage to turns that carry a form descriptor. For questions about the product itself, the always-offered `read_user_guide` tool (`orchestrator/guide_tools.py`) returns any `guide/prebuilt/<id>.json` section; its description lists the section ids, fixed at import, so it never changes the cached tools prefix.\n\n**Take the lead** (`orchestrator.take_the_lead`) is how the Executive acts on its own, behind one gate. It has two switches, both off until someone turns them on. **As the Executive** is the principal's alone (`PUT /take-the-lead`). With it on, the unattended passes (reflection, research, the scheduler's proactive trigger) keep the acting tools solo mode otherwise withholds from them: booking, starting workflows, and messaging teammates rather than only the principal. `schedule_tools.unattended_toolkit` and the chat loop's unattended path then wrap every acting tool (`ACTING_TOOLS`, plus any MCP tool whose name sends, creates, updates, deletes or shares) in `gated_handlers` / `gated_call_tool`. The principal-only tools (`UNATTENDED_WITHHELD_TOOLS`) stay withheld, and every guard inside a handler still applies, including department approval levels (a booking in a department on Proposes is still a proposal) and the contact-egress rule. New mail or chat (any turn with an `origin_channel`) wakes the existing reflection early (`take_the_lead.wake`): one run waiting at most, five minutes after the first message and no more than every 30 minutes, as a one-off `take_the_lead_wake` scheduled action. Pause holds it like any scheduled work. **As you** is the Handle it for me dial's top step (delegation section).\n\n**The gate** is plain code (`take_the_lead.check`). Six kinds of action wait for a yes first, each with an Ask first switch, all on to start. The switch is the decision ledger's own mode for the class `take_the_lead_action:<kind>` (`propose` asks first, the default; `auto_execute` doesn't). The kinds are money, contracts and legal, hiring and people decisions, deleting/cancelling/sharing, someone new (an address on neither the roster nor the contacts) and big sends (a broadcast or department message, or more than five people). On top of these, the company (set by the principal) and each person (for their own As you) add structured rules: a person or address, a domain, words, or an amount at or above a figure. Added rules always hold, and an error reading them holds the action. The one thing the gate lets straight through is a direct message (`message_person` or a Slack, Telegram or Discord DM) whose only recipient is the principal (`_only_to_principal`): they are who it would wait for, so a card would only ask them to approve a message to themselves. Anything that reaches or acts on anyone else is checked, and a roster that can't be read is not the principal. A held action becomes a `take_the_lead_action` decision instance with a companion `DECISION_ALERT_SOURCE` card on Today, routed to the first teammate holding one of its People approval ranges (`take_the_lead.approval_ranges`, plain code, closest fit first: money to `spend_lt_2k`, `spend_lt_10k` or `spend_gt_10k` by its largest amount, with a higher range able to approve a lower one and no amount read as large, or to `customer_credit` for credit and payment terms; vendor contracts to `vendor_onboarding` then `legal_sign`, other contracts to `legal_sign`; hiring to `hiring_signoff`; anything else about the board or investors to `board_comms`; in Personal as in team mode), or else the principal through the authority gate's fallback (`departments.authority._route_proposal`). Deletes and shares always go to the principal. Only the principal and that approver see or resolve it (`DecisionClassSpec.principal_and_approver`). Approving it claims the row and carries out the exact call once (`take_the_lead.carry_out`: the unwrapped handler, or the gateway's `call_tool`, on an unattended session, with contact egress only when the principal approves). Declining drops it. Everything done or held is a row in `take_the_lead_log`. What it did and what was approved show in Recent activity (`today._build_activity`, kind `took_the_lead`), named but never quoted, since that feed is the team's. The held card's own `decision_resolved` row is left out because its payload quotes the message. `GET /today/activity` also adds the caller's own Handle it sends (kind `sent_as_you`, actor `you`), which are never in the shared feed.\n\n**In training.** The As the Executive switch has three positions: Off, **In training** and On (`take_the_lead.training`). In training, an action the gate would let through still waits, as a `training` hold for the principal alone, unless the principal has allowed that action with that person or thing (`take_the_lead.allowance`: \"Message Priya Nair\" covers a message to her on any channel; \"Book meetings with Priya Nair and Sam Lee\" meetings with exactly them; a workflow by id; a connected tool by name and the file it names, never one that reaches people such as a mail send or a share). The card carries Approve with a **Do this on its own next time** tick (`allow`), **Edit** (the message, a meeting's title, times or description, a task and its due date: `take_the_lead.apply_edits`) and Decline. Allowing is the principal's alone and needs a request the API can tie to them, as turning it on does; the six kinds and the added rules hold whatever is allowed. An edit approved with \"Do it like this next time\" is kept as an example (only the edited text, the principal's own words), and `take_the_lead.learned_note` hands the examples and the fact it's in training to the reflection and research passes in their user turn. After `SUGGEST_AFTER` (3) unchanged approvals of the same action in 30 days, Settings suggests allowing it. What it's learned is one list keyed by feature (`take_the_lead_allowed.feature`), so other features that learn from approvals can join it; each keeps its own switch. Act as me is the second (`act_as_me`, per person, `delegation.training`), with a training switch per setting (Replies, Follow-ups, Suggested actions, Drafts): its rows carry the person's id and the setting in their key, are listed and removed by that person alone, count toward their own cap, and never reach `learned_note`.\n\n**AI disclosure.** Nothing the Executive sends to people outside the web chat goes without saying it is an AI: plain code adds the line in the MCP gateway (mail) and at each chat send (Slack, Telegram, Discord, Google Chat replies, its DMs and room posts), never the prompt, so the persona and the cached system blocks are unchanged and a reply the model writes can't drop it. Replies the inbox watcher writes as a person carry their own line by that person's switches. See Integrations → **AI disclosure** and Act as me → **AI line on replies**.\n\n**App control from chat** (`orchestrator/app_control.py`, catalog in `orchestrator/app_control_catalog.py`). Three tools in the `app` tool group let the Executive do what the app's buttons do when the person asks (\"pin this\", \"delete that chat\", \"turn off Handle it for me\"): `find_in_app(kind, query)` lists what the speaker may change (their chats, topics, documents, workflow runs, settings; files, memory, departments, skills and workflows for the principal), `do_in_app(action, id, value)` does one action (`id: \"this\"` is the chat they are in), and `open_in_app(screen)` returns an in-app link. Each action calls the API route its button calls, as a function, with a request naming the turn's verified speaker (`correction_tools._speaker`; an email or unattended turn gets a refusal), so the route's own checks decide what this person may do. Actions whose route checks nothing about the caller (their page is the principal's) are `principal_only`. Deletes, shares, turning on Act as me, Handle it for me or Take the lead, removing a Take the lead rule, clearing every card and approving a workflow's waiting step never run in the turn: they become an `app_action` decision card, private to the speaker, shown under the reply; the tap runs the action with the tapper's own request through `POST /decisions/{id}/approve` (claimed once; a card older than a day is turned down). After a turn reads outside text or the speaker's mail, every action becomes a card. The tools are withheld from unattended and AI-app turns, and `do_in_app` from a shared chat someone else spoke in. Snoozing a card uses `ack_alert`'s trust rule: only ids the server showed or found this turn. Extensions add kinds, actions and screens with `register_kind` / `register_action` / `register_screen` at startup. The person's notes stay with `recall_history`; account, sign-in and billing stay on their screens.",+ "markdown": "The Executive (`openexecutive.orchestrator.executive.Executive`) is a single coherent persona that fans out to specialist sub-agents via Anthropic tool use. Each specialist subclasses `openexecutive.agents.base.BaseAgent`, carrying a `name`, `domain`, and `model`, and exposes `analyze(...)`, which wraps the question in user-turn context and calls Claude through the per-call provider. That context is per call and never cached: `<company_stage>` (the stage from the company profile — specialists never see the profile itself), in solo mode `<principal_role>` (what kind of principal the advice is for, in plain words, with their title and remit — `router.principal_role_context`), `<department_memory>`, `<past_decisions>`, `<failure_cases>` (the shipped case studies, and under them the workspace's own lessons that match the call, `memory/lessons.py`), `<relevant_knowledge>` and `<conversation_context>`, outermost first. The specialist's own system prompt is a static domain constant under one `cache_control` block, so it stays byte-identical across companies and profile edits. A chat turn takes the stage from its session's profile (the one the Executive reasons over, and the one an eval scenario injects); workflow steps and the MCP server's `consult_specialist` read it fresh from the profile on disk. The role tag follows the turn's mode: a chat turn pins the role on its session with the mode (`pin_turn_principal_role`: the session's `principal_role` override, else the workspace settings) and sends it in solo and nothing in team, so the org block, every specialist in the turn and any workflow the turn starts see the same role; callers with no chat turn read it fresh (`load_principal_role`: the current session's mode and role, else the workspace's), and a failed read sends no tag rather than failing the consult. The CFO, CSO and sales prompts key their advice on it: for a bootstrapped, self-funded or solo company they drop venture fundraising benchmarks and work from cash, and raise only if the founder wants to.\n\nRouting uses one `consult_specialist` tool. When to call it is spelled out in the persona's Domain Expertise section (\"Going deeper\") and in the tool's own description: any question that asks for judgment in a specialist's area (a decision, plan, analysis, negotiation, hire, contract term, pricing or fundraising move, board message, or a department's own targets and stance, since the specialist holds that department's memory) is consulted before the answer, several at once across areas; lookups in mail, calendar or documents, scheduling, status checks, routine drafting and questions about how the Executive works are answered without one. Chat eval scenarios hold this: `required_routing` names specialists the turn must consult and `no_routing: true` says it must consult none, and the runner fails a scenario that misses either (`evals.scenarios.routing_check`). Beside it the Executive carries two skill tools, `search_skills` and `load_skill` (skills are shown to users as **playbooks**, and every playbook belongs to a workflow: one that follows it, or its own Quick workflow). Chat finds and reads playbooks but never creates or changes one: work worth repeating is offered as a workflow through `draft_workflow`, with the method written into its steps, because workflows read playbooks at run time (approved scheduled ones included) and this loop also runs on inbound email and chat, so only a person can change what a workflow follows. A hidden built-in is invisible to search and load; each `search_skills` hit lists the `workflows` that follow it, Quick ones included, and the persona routes \"the finished deliverable\" to that workflow and quick or partial asks to the playbook inline; Tool calls made while proactive work is tagged (scheduler dispatch, reflection, research, alert review) feed the Attunement outcome ledger. Goals have three tools: `list_department_goals`, `update_department_goal` (progress the user reports) and `create_goal` (a new goal the principal states, filed under an area by slug or title and creating the area when none matches — see Org). `create_goal` runs only for the principal on a verified surface, and no unattended run is offered it: the tools in `schedule_tools.UNATTENDED_WITHHELD_TOOLS` are dropped from reflection's and research's toolkits by `unattended_toolkit`, and from the chat loop when its Session is marked `unattended` — the scheduler's PROACTIVE TRIGGER run, whose prompt quotes stored intent text; a call the model emits anyway gets an error tool_result. `record_decision_outcome` records how a past decision turned out — the weekly review lists decisions older than 30 days with no outcome as `[decision N]` — writing `decisions.outcome` (`episodic.update_decision`) and auditing the rationale; it has `create_goal`'s gate (the principal on a verified surface only; everyone else gets an audited refusal) and is in `UNATTENDED_WITHHELD_TOOLS` too. A missing or unknown id returns the decisions still waiting on an outcome, so the model can retry with the right one. Its schema is static; the solo persona has one line on it, and team mode learns it from the tool description, so the team persona is byte-identical. Attunement adds `list_open_loops` (what a person owes — everyone's for the principal only in a private conversation, otherwise the caller's own) and `close_open_loop`, which only closes a loop when the turn's resolved speaker is the principal or the loop's owner — an unrostered sender, a signed-in web user not on the roster, and unattended sessions are refused. `assign_open_loop` opens one on request (\"ask Ben to send me the Q3 numbers by Friday\") for the principal or a rostered teammate on a verified surface — never an email, a turn private to the principal or an unattended run, which are also withheld the tool — and only when `task` appears verbatim in what the speaker typed this turn, at most 5 per turn; that turn's extraction pass then opens no second loop for the same owner. The owner is told once on their own channel when someone else assigns it, and the result's `notified` says how that went, so the Executive reports it rather than messaging them as well. The roster write tools (`upsert_person`, `archive_person`, `set_department_head`) are the principal's alone, and only on a surface that verified the speaker — the web chat, the principal's own Slack or Discord, a channel whose adapter verified the sender itself (`Session.speaker_verified`, with the `caller_person_id` it resolved, never on a turn an email started), or a private Telegram chat when `TELEGRAM_WEBHOOK_SECRET` is set to a value Telegram can send (Telegram identifies the chat rather than the sender, and without that secret its webhook accepts anyone's update). A teammate, an email (even from the principal's address), Google Chat, the CLI, the MCP server and unattended runs get `status: refused` (audited, and no action chip), because a team member's People row is also the web sign-in and outbound-email allow-list. `upsert_person` takes `kind` (`team` / `contact`; a new person defaults to `contact` in solo mode, an update keeps the current kind) and `email_aliases` (other addresses the person writes from — the full list, replacing the current one; they match mail and may be emailed but never sign in, and one already on someone else is refused), and after every write any pending roster request whose sender now matches someone is closed and its held messages replayed. `resolve_roster_request` answers a **roster request** — someone not on the People list who wrote in and is waiting (see Org): `approve` (a new person; `kind` defaults to team for a company-domain sender, else contact, when the principal did not say), `link` (someone already listed, by `person_id`; `replace_channel_id` only when the principal said so) or `decline`. It has the roster tools' gate (the principal on a verified surface; audited refusal otherwise), is in `UNATTENDED_WITHHELD_TOOLS`, and answers only an id in `Session.trusted_roster_request_ids` — the ids the server printed that turn in a `<roster_requests>` block (`briefing.context.format_roster_requests_for_prompt`, up to 10 pending requests), which `render_and_trust` appends to the `<briefing>` digest on the principal's own verified turn only and records on the session in the same step, so an id the model invents or reads out of a message adds nobody. Each line is server-derived — id, channel, address or account id, the sender's name sanitised and marked unverified, a company-domain hint, the message count — and never what they wrote. The block rides in the **user turn**, like the digest, so the cached system blocks are unchanged; the tool definition is static, and the new tool only extends the sorted tool list. The web chat route resets the trusted set every turn, and a resolved id is removed from it. and `list_people` returns both kinds with `kind` — contacts only on the principal's own verified turn (they are private to the principal; anyone else gets the team as if no contact existed), each with what is remembered about them from the principal's own mail and meetings (`what_you_deal_with`, `topics`, `recent_activity`, from `people.contact_notes`) so the Executive can suggest who to ask, and their contact details (`details`: phones, links, address), both on that turn only — while `lookup_person` searches the team only. `ask_about_person` on one of the principal's contacts, on the principal's own turn, returns that contact note instead of peer memory (contacts never chat, so peer memory has nothing on them); otherwise it reads a person's peer memory fully only on their own verified turn; anyone else's answers only to a verified principal or teammate, and only when that person turned on \"Share my work style with the team\", framed to how they work and never their view of a colleague, and otherwise returns the same \"hasn't chosen to share\" result whether or not anything is known. A teammate's own mail with the Executive is theirs the same way: every mailbox call the model makes through the MCP gateway (Gmail or Outlook, by `call_tool`, from a script, or a workflow step run from chat) passes through `orchestrator.teammate_mail`. A search excludes every other teammate's address server-side, except on mail the reader is also on, and is filtered again entry by entry; any other call that names a message or thread (read, attachment, reply, forward, label) must use an id a filtered search returned to the same reader in this conversation (`Session.mail_ids_seen`) or the inbound email the turn answers; reads are filtered message by message, so a message naming another teammate is left out unless the reader is in its From / To / Cc; a reply that would quote such a thread, or an Outlook forward, reply or attachment read of such a message, is refused; and a conversation in which a teammate other than the principal was shown their own mail becomes theirs alone, as Act as me's do. Outside mail and the reader's own are untouched; an email turn reads as its sender only when Gmail authenticated it, and any other unverified reader sees no teammate's mail, a background run the model drives reads as the principal and a workflow step outside any session as nobody, an Outlook inbox rule that forwards or names another teammate is refused, and the poller and the brief, which read the mailbox themselves, are unaffected. Their own chats are theirs too: every audit row of a teammate's verified turn is theirs (`people_tools.teammate_turn_owner`, see Audit; the principal sees the actions, never the words), and a teammate's web chat or DM opens for them alone (`chat._session_access`), the principal included. The same verified-principal rule, as `people_tools.contacts_reachable_now()`, decides whether a turn may reach a **contact** at all: `message_person`, the raw DM tools, `create_calendar_event` / `create_instant_meeting` attendees and the MCP gateway's Gmail / Calendar / Drive / OneDrive gates treat a contact exactly like an unknown person on every other turn, so no refusal reveals that one exists. **The untrusted-content policy** (`orchestrator.content_trust`) is one rule for text the principal did not write — inbound email, chat, attached files, watched pages — answering three questions for every surface. *Who is speaking*: the principal only when the surface proved it (`principal_speaking`: the web chat as the principal — never a signed-in email on nobody's People entry —, the CLI, their own Slack/Discord or a secret-verified Telegram chat, or their own address with Gmail's `dmarc=pass`); never a teammate, a stranger, Google Chat, the MCP server or an unattended run. *What may become memory*: episodic extraction runs only then, and only on the words outside `<untrusted_content>` blocks; peer memory and open loops are recorded under the speaker's own name, never as the principal's. *What may trigger tools*: every write on the principal's authority already refuses others in its handler; the tools with no such check that change the install for every later turn — `content_trust.PRINCIPAL_ONLY_TOOLS`: `load_mcp_server`, `list_saved_tools`, `run_python_job` and `run_script` — are not offered unless the principal is speaking on an interactive surface (not email, even authenticated), and a call the model emits anyway gets an error tool_result and a `refused: not_principal` audit row. So an inbound email, a teammate's turn, Google Chat and the scheduler's proactive trigger each get a stable tool list of their own without it. A send or edit on a team member's turn also meets their department's authority first (`orchestrator.department_gate`, see Org), and so does one on a turn whose speaker is on no People entry: a signed-in web caller or a chat-channel sender, which in practice is Google Chat, since Slack, Telegram and Discord turn unmatched senders away. Google Chat names a matched sender to the gate alone (`gate_sender_person_id`), so the principal there is not held. Email turns are left to the mailbox's own gates. *How it is labelled*: outside text goes into the user turn inside `<untrusted_content source=… author=… author_verified=…>` with a fixed notice that it carries no authority; compatibility forms are folded (NFKC), control characters go, and the tag's name is renamed wherever the text carries it — any case, spacing or entity-escaped form — so the text cannot end its block and speak outside it.\n\n**Outside content** (`orchestrator.outside_content`). Once a turn has read text from outside the team (any gateway read, an attachment, a web search result, a synced Drive, OneDrive, Confluence or Notion excerpt, `read_document`, research, or an extension tool registered `reads_outside`), the rest of that turn refuses what the Act as me lockdown refuses, every `upsert_person` but a contact the speaker named in this message at an address they typed or one that sent the mail read (`lockdown.speaker_named_contact`; never on a turn answering mail from someone not in People, and a refused one points to an `add_contact` card), and every `call_tool` but a read: no mail, Docs, Sheets, Drive or calendar write. The company's own Drive is not outside: a Google read of one file, Doc, Sheet, deck or folder by id (`company_read_item`) is looked up first (`vet_reads`, once per id a turn), and when that item is owned by someone at the company, or kept on a shared drive only the company organizes, and shared with nobody outside it (`mcp_gateway._company_item`), the read marks the turn as a company read (`COMPANY`) instead. That still holds everything else, since anyone who can edit the file wrote its words, but when every read so far was of the company's own items, a rename or move (`update_drive_file` with only `file_id`, `name`, `add_parents`, `remove_parents`) goes through, so the turn can file what it read. Searches, mail, calendar and forms mark it as usual. A refused rename or move the person asked for can go on a `drive_files` approval card instead. A round that reads also holds that round's other such calls; the read itself runs. Scripts are checked call by call. The offered tools never change, so the cached prefix stays the same, and the hold ends with the turn. Mail from someone not in People (a contact or a stranger) counts as outside content from the start of its turn (`Session.sender_off_team`), so the whole turn is held, with no direct messages or watchlist removals either; it can still raise an alert, private to the principal, which is how that mail reaches them. Someone added to People in a turn (or given a new address or chat account) is not emailed, messaged, invited or shared with until the next one. What a turn held is remembered for the conversation's next turn (`note_held`, in process, for an hour). It is kept per conversation and speaker. When that speaker's next message is a go (a short yes such as \"go\" or \"yes, do it\", or one opening with go, go ahead, do it or proceed as its own clause: `GO_RE`), the first round that reads has its reads refused once (`defers_read`, with an error that names no tool and asks for nothing new), so the steps the person just approved run before a new read holds them again; from the next round reads run as usual. It runs only what the model calls in that turn, under every usual check. A shared chat, and a turn answering mail from someone not in People, remember nothing. In the web chat the go is a tap: while it waits, the reply carries **Approve** and **Not now** buttons naming the held steps in plain words (`pending_held`, `step_words`: only names it knows; any other gateway tool shows its app's generic label, since the model's tool name can be steered by what it read), and Approve sends \"Go ahead\" as the person's own next message. The refusal on a web turn (`approve_hint`, set from `Session.from_web_chat`) tells the model to point them at Approve rather than ask them to say go; other surfaces still ask for a typed go.\n\nHow much a read holds is the owner's setting, **When I read outside content** (`orchestrator.outside_mode`, one row in `outside_content_mode`, changed only by the principal through `GET`/`PUT /executive/outside-content` on Settings → Your Executive), fixed per turn in `begin_turn` (`TurnOutside.mode`). **Ask every time** (`ask`) is the hold above. **Only for outsiders** (`outsiders`, the default) skips what came from the team: a Gmail search that shows only ids and links, a Gmail message read where every message's sender is a teammate (not a contact) and Gmail authenticated it (`fact_confirmation.sender_authenticated` on a raw read of each id, at most 10) and no other address appears, and a calendar read naming only the team where the Executive itself organized every event (a teammate's invite counts as outside, since its organizer is whatever the mail that brought it says) (Google's `get_events` only when `detailed` or for one event, since the plain listing doesn't say who made each event; Microsoft 365's calendar reads) (`mark_read` / `from_the_team`, checked on the result; an `@` the address pattern can't read counts as a stranger). Attachments and synced folders (a file carries whoever wrote it, not whoever sent it), web pages, fetched documents, vendor apps, Outlook mail and every other gateway read still hold. **Don't ask** (`trust`) applies only on the principal's own turn on an interactive surface that isn't email or a shared chat (`content_trust.principal_only_withheld` empty): no read marks the turn, and reading their own mail holds nothing either (the Act as me lockdown and `app_control` are skipped via `trusted_turn`); anyone else's turn gets `outsiders`. Whatever the setting, a shared chat, another AI app's turn and mail from someone not in People are `ask`, sends still go only to people on People, and someone added this turn is reached no earlier than the next one. An unreadable setting is `ask`.\n\nA turn about the principal's private mail (`Session.private_to_principal`) reaches the principal only. It is neither offered nor allowed to run `schedule_tools.PRIVATE_TURN_WITHHELD_TOOLS` — everything that reaches someone else, publishes where the team reads, or starts work outside the turn: the broadcasts, meeting bookings and cancellations, `run_workflow` and research, follow-ups and workflow nudges, watches, artifacts, goal updates, saved workflows, skill drafts, and `load_mcp_server` (it connects to any URL, which can carry the turn's content). Through `search_tools` and `call_tool` it reaches only `schedule_tools.PRIVATE_TURN_MCP_TOOLS`: Google Workspace reads (mail, calendar, and Drive, Docs and Sheets search and file reads, with their Microsoft 365 twins; OneDrive's `download-bytes` only for a file's content) and the Gmail send and draft the gateway narrows to the principal. A search shows it nothing else, and a call to anything else (Google Chat, a Docs or Drive write, `manage_event`, another server) is refused (`schedule_tools.private_turn_withholds`). A call the model emits anyway, MCP tools included, gets an error tool_result and writes an audit row, and every audit row the turn writes is the principal's alone.\n\nA turn another AI app asks over the MCP server (`Session.from_mcp_client`, set by `ask_executive`) reads and advises but never acts, unless the request may act (below). It is neither offered nor allowed to run `schedule_tools.MCP_CLIENT_WITHHELD_TOOLS`: the private-turn and unattended sets, plus every direct message (`message_person`, the Slack, Discord and Telegram DMs), alert acks and new alerts, brief edits, closing open loops, roster and department-head writes, watchlist changes, workflow drafts and form proposals. It gets no Act as me, `recall_history` or extension tools, and of the gateway tools only `schedule_tools.MCP_CLIENT_MCP_TOOLS`: the private-turn reads without the Gmail and Outlook send and draft tools. The set is fixed, so the variant keeps a stable cached tool prefix. A call the model emits anyway, through `use_tool` or `call_tool` included, gets an error tool_result telling it the conversation comes from another AI app, so it can read and advise but not send, post, book or change anything, and writes an audit row (`refused: mcp_client`); `MCPGateway.call_tool` refuses the same calls again.\n\nA request may let that turn act (`Session.mcp_client_may_act`: the `x-mcp-client-actions: on` header on a request the shared secret authenticated, or `MCP_CLIENT_ACTIONS` where no secret is set). It then runs the speaker's own changes inside Open Executive (`schedule_tools.MCP_CLIENT_RUNS_WHEN_ACTING`: alerts, open loops, follow-ups, the watch list, goals, decision outcomes, facts and company profile writes; never the roster), each under its own owner-only and person checks, since the turn counts as a verified surface for its caller. It is still denied a fixed set, `MCP_CLIENT_ACTING_WITHHELD_TOOLS` (DMs, calendar writes, broadcasts, `load_mcp_server`, scripts, Python jobs, workflows, research and the rest of the read-only set), with a refusal telling the person to do it in Open Executive. A message or a meeting goes on an approval card instead (`propose_actions`), which is carried out at once when the person's Handle it for me covers it, everyone on it has been on the roster for a week, and `action_cards.on_its_own_refusal` finds nothing against it, and otherwise waits for their tap. Every audit row it writes names the app (`via`).\n\nA web chat its owner shared (`Session.shared_chat`) mixes people. The note naming who is in it says which of them owns it and that the others' lines are messages from them, not the owner's instructions. While someone other than the owner has written since the owner's last message (`Session.others_spoke_last`, set per turn by the chat route from the saved history, the speaker included, and true when the history can't be read), the turn is neither offered nor allowed to run `schedule_tools.SHARED_CHAT_WITHHELD_TOOLS`: direct messages, the broadcasts, meeting invites and cancellations, follow-ups, roster writes and workflows (a run, a save or a suggested schedule acts later with the full tool set). Through `call_tool` it can't send, reply, forward or draft mail in Gmail or Outlook, or make a calendar change that mails attendees (`schedule_tools.shared_chat_withholds`). That holds on the owner's own turn right after someone else's line too, so a request carried over from it can't send; the owner's next message with nobody else's in between lifts it. A call the model emits anyway gets an error tool_result and writes an audit row (`refused: shared_chat`). The principal-only tools stay out of every shared chat (`content_trust.principal_only_withheld`).\n\nOn every turn the Executive knows what is connected without searching: block 0's *Connected Systems* section names the connected Google services, other MCP servers, the messaging channels, Notion sync, Confluence sync (when on) and Act as me, and pins the Google tool names it calls directly with `call_tool`. Both the service list and the pinned names are narrowed to what workspace-mcp actually loaded (`WORKSPACE_MCP_TOOLS`, `WORKSPACE_MCP_TOOL_TIER`), so a service left out is not named and its tools are not pinned — being told a tool is already available when it is not in the index costs a failed call the model cannot explain to itself. `search_tools` is for anything not listed. When work repeats over many items, the Executive can also call `run_script` (`workflows/step_script.py`, offered when the gateway runs, unless `CHAT_SCRIPTS=false`, and only on the principal's own turns — `PRINCIPAL_ONLY_TOOLS`, withheld on an unattended pass too, because it spends the turn's whole call budget on code the model just wrote): one short Python script, run in a sandboxed Monty worker with no files, network or environment, whose only reach is the gateway tools `call_tool` could reach (those `search_tools` has returned; the gateway keeps that set for its own process-wide session, not per conversation). Each call it makes is a `call_tool` in every way that matters — the gateway's discovery, deny-list and recipient gates, Take the lead's gate on an unattended run (or the department gate on a team member's turn, see Org), and the same action chip and `tool_invocation` row (marked `via: run_script`, `ok: false` when it raised), as each call happens; the script itself gets a `kind: script` row with its source and result. Its wall clock is half the turn's deadline, so a long script ends with the list of calls it made rather than being cut off with the turn, and a stopped turn closes it and its worker. A turn private to the principal is not offered it, and the Act as me lockdown withholds it whole. Besides gateway tools, a chat script may call a fixed list of the Executive's own tools (`step_script.CHAT_OWN_TOOLS`: reads, alerts, open loops, goals, the watch list, follow-ups, `message_person`, `upsert_person` — nothing that broadcasts, books calendars, edits memory or skills, or starts workflows or research), each only when the turn is offered it and through the turn's own handler (Take the lead's gate included), with the same chip and a `kind: skill`, `via: run_script` audit row. A direct call needs a `tool_use` each, so a turn's scripts share a budget instead: `CHAT_SCRIPT_MAX_CALLS` (default 200) calls in all, and per-turn caps on own tools that could otherwise act in bulk (`CHAT_OWN_TOOL_CAPS`: `message_person` 10 — messaging the whole roster would be a broadcast — `upsert_person` 25, `create_alert` 50); a refused call tells the model to say what is left. A call stopped mid-flight by the script's clock or the turn still gets a `cancelled` audit row, since it may have run. Because every round re-reads the conversation, the loop also nudges at the moment of choice: after a round whose results include a list of five or more items (`step_script.lists_many`), with `run_script` on offer and not used that round, it adds the constant `FANOUT_HINT` as a text block in that round's user message (never a cached block), once a turn, pointing the per-item work at one built tool. For library work on files (PDFs, spreadsheets, charts, images) the chat also has `run_python_job` (see Workflows → Python jobs): Python with a fixed library set in a Deno-isolated WebAssembly sandbox, fed Knowledge files by name, the message's own attachments as sent, or text it already holds, and able to keep a job that worked as a Custom tool, with result files linked for the principal to download; it is offered only on the principal's own turns and only where the sandbox is installed (not offered otherwise, so the tool prefix stays fixed per process). Agent Activity shows each run as one \"Built tool\" card (debug event `script_run`: ok, the calls it made, how many, how long, and the kept or reused tool's name). The persona's tool-building section is present only when chat scripts can actually run (the setting and Monty installed). `TOOL_BUILDING_ADDENDUM` in the cached persona tells the Executive when to build a tool — a job that looks something up and then checks or acts on each result, paged results, more than about ten items, or an answer that needs three or more reads (find, open, read; the tool returns only what answers the question, not whole documents, and only what is about to be edited or quoted at length is read in full), while a handful of items already named go as direct calls in one turn — and never to say it has none, and to talk about it as building a tool or doing it in one go. On the principal's own verified, interactive turns, a script that works can be saved as a named tool (`save_as`; elsewhere the script runs but is not saved) and run again with `run_script(tool=…, inputs=…)`; `list_saved_tools` (a constant definition, in `PRINCIPAL_ONLY_TOOLS` with `load_mcp_server`, so offered only while the principal is speaking on a verified, interactive surface; `run_script(tool=…)` is refused on any other turn) lists them, and each call a saved tool makes is a gateway call like any other (see Workflows → Saved tools). The tool's definition is a constant, so the cached tool prefix does not move. Multiple parallel `tool_use` blocks in a single assistant turn fan out concurrently through `router.route_parallel`, each running its own domain-filtered ChromaDB query (3 built-in and 2 company chunks by default, see the RAG section) before the specialist sees the question. The fan-out is failure-isolated per call: a specialist that raises or returns no text becomes an `UNAVAILABLE` tool_result (`router.specialist_unavailable_result`) instead of failing the whole turn, and its area — \"finance\", \"legal\", never the specialist's name — is reported to the user under the web reply, or by the Executive in its reply on other channels, unless another call to it answered that turn. A failed specialist does not count as consulted, so it takes no committee seat. Cancellation (Stop) still propagates. The Executive then synthesizes the returned analyses into one reply.\n\n`remember_fact`, `forget_fact` and `update_company_profile` (`orchestrator/fact_tools.py`; the last two sit in the `memory` group of `open_tools`, since they are rarely needed) make a correction stick: the first two keep or retire a standing fact in `memory.facts`, and the third changes one field of the company profile. They share `record_decision_outcome`'s gate (the principal on a verified surface, an audited refusal for everyone else) and are in `UNATTENDED_WITHHELD_TOOLS`. `remember_fact` alone also admits a rostered teammate on a verified surface (web signed in, their own Slack or Discord): the fact is stored attributed (`recorded_by_role=\"teammate\"`, rendered `(per <name>)`), and is held as a `proposed` row (`awaiting_approval`) until the principal approves it, unless the principal marked that teammate trusted (\"needs my approval\" off; on by default) — and even then when it would replace the principal's fact. All three are also in `PRIVATE_TURN_WITHHELD_TOOLS`, since what they write or retire is read on everyone's turns. On the principal's own authenticated email (primary address, with Gmail's own `Authentication-Results` reporting `dmarc=pass`, not forwarded) the three tools check the request as usual, then hold it as `awaiting_confirmation`. It applies only when the principal replies to a one-time token emailed to their own address (`integrations.fact_confirmation`); the split into a check step and an `apply_*` step lets chat and the confirmed reply share one apply path. Each write also needs a `source_quote` found in the speaker's own words this turn. The active facts reach every specialist as a `<standing_facts>` user-turn tag: `route_to_specialist` reads them (None → the store, \"\" → none) for chat consults and workflow steps alike, and `BaseAgent.analyze` places the tag next to `<past_decisions>`, never in the cached specialist system prompt.\n\nThe ten entries in `SPECIALIST_REGISTRY` (`openexecutive.orchestrator.router`):\n\n| Key | Agent | Domain |\n| --- | --- | --- |\n| `cso` | `StrategyAgent` | strategy |\n| `cfo` | `FinanceAgent` | finance |\n| `chro` | `HRAgent` | hr |\n| `gc` | `LegalAgent` | legal |\n| `coo` | `OperationsAgent` | operations |\n| `cmo` | `MarketingAgent` | marketing |\n| `cpo` | `ProductAgent` | product |\n| `sales` | `SalesAgent` | sales |\n| `board_comms` | `BoardCommsAgent` | board |\n| `triage` | `TriageAgent` | triage |\n\nModels: specialists default to `claude-sonnet-5-5`; the deep-reasoning tier (CSO, CFO, GC, Board) runs on `DEEP_REASONING_MODEL`, `claude-sonnet-5-5` by default, with adaptive thinking; routing/fast classification uses `claude-haiku-5-5`. Requests are fitted to the model's generation on the way out (`feature_gate.fit_claude_generation`): 5-generation Claude models never receive `temperature` / `top_p` / `top_k`, and Haiku 5.5, which thinks unless told otherwise, gets thinking turned off when the caller asked for none. The Council's **Deep reasoning** toggle is honoured on Claude Opus/Sonnet (direct or via OpenRouter) and on any OpenRouter model whose live catalog entry supports `reasoning` (today all OpenAI, Gemini, DeepSeek and Grok entries; not Llama) — the request translator turns Anthropic adaptive thinking + effort into OpenRouter's `reasoning` parameter. It is a silent no-op on models that can't reason, and disabled for Haiku, which rejects adaptive thinking. A PDF `document` block is translated too: it becomes an OpenAI `file` part, and OpenRouter requests carrying one add the `file-parser` plugin (`native` for models whose catalog entry lists `file` input, else `PDF_OPENROUTER_ENGINE`, default `mistral-ocr`). `FeatureSpec.supports_pdf_input` gates it per model — off for a local server unless `LOCAL_PDF_INPUT` — and a gated-off PDF becomes a text note rather than vanishing. The Council picks a model in two steps: a **Provider** dropdown (Anthropic, OpenAI, Google, Meta, DeepSeek, xAI, Local, …), then a **Model** dropdown limited to that provider, with readable names (\"Claude Opus 5.5\") and the raw id and serving route shown underneath. The Anthropic group combines the direct Claude models (`claude-fable-5-1`, `claude-opus-5-5`, `claude-opus-5`, `claude-sonnet-5-5`, `claude-sonnet-5`, `claude-haiku-5-5`, `claude-haiku-4-5`) with any Claude generation that exists only on OpenRouter, and never lists a model twice. With `OPENROUTER_ENABLED`, the non-Anthropic entries come from OpenRouter's live `/models` catalog (fetched at startup, refreshed in the background, filtered to tool-capable paid text models from an allowlisted vendor set), falling back to a built-in snapshot if the fetch fails. Every Council agent also takes **additional instructions** (`agent_overrides.instructions`): text appended after its prompt by `agents.overrides.append_instructions` inside an `<additional_instructions>` block. `BaseAgent.base_system_prompt()` is the built-in prompt or its full replacement (what the Council's editor shows); `effective_system_prompt()` is that plus the instructions (what every call sends), and the Executive gets the same text through `build_system_blocks(persona_instructions=…)`. Unlike a replaced prompt, instructions leave the agent on its built-in prompt, so prompt improvements keep reaching it. A **Quality** choice above the agents (`agents.presets`) sets every agent's model and deep reasoning at once: **Fast** puts every agent on a small model (`claude-haiku-5-5`, else `claude-haiku-4-5`) with deep reasoning off, **Balanced** returns each agent to its built-in defaults, and **Thorough** puts the Executive and the domain specialists on the largest allowed model (`claude-opus-5-5`, else `claude-opus-5`) with deep reasoning on where the model supports it, leaving triage and the helpers below on their defaults. Presets change quality only; they never cap spend. Each preset resolves against the install's model allowlist at request time and is offered only when one of its models is allowed. Applying one writes ordinary per-agent overrides (model and deep reasoning only), so history, rollback and later per-agent edits keep working; an agent changed on its own afterwards shows as **Custom**. Each Council agent has a **visibility**: `core` for the Executive and the domain specialists (`BaseAgent.visibility = \"core\"`, which a new specialist must declare) and `internal` (the default) for triage and the helper agents below. Thorough moves only core agents. The Council opens in a **simple view**: the Quality choice, the voice picker, and only the core agents, each with its additional instructions and a model badge. **Show all agents** lists the internal ones too, and **Advanced** opens the full page (the browser remembers the choice). Eight Council-only agents are model/prompt-overridable in the UI but NOT callable via `consult_specialist`, because they are deliberately kept out of `SPECIALIST_REGISTRY`: `quality_judge` (committee review), `utility_fast` (gates, titles, decision parsing), `research` (the `executive_research` fan-out), `fixture_generator` (invents demo companies), `engagement_intake` (models a real client from intake material), `onboarding_interviewer` (interviews a new user about their own company and drafts its profile, people and departments), `workflow_designer` (runs the conversational New workflow wizard on `/jobs/new`: searches the available tools, asks clarifying questions, then drafts a validated dynamic-workflow definition for the user to review), and `workflow_actor` (runs a workflow's `action` steps: a bounded tool-use loop limited to the tools the user approved for that step). `fixture_generator`, `engagement_intake` and `onboarding_interviewer` share one shape — a forced tool call producing a whole company, validated for referential integrity, with a single repair retry — but opposite epistemics: the fixture author invents, the other two extract and leave unstated facts empty rather than guessing. When the workspace is in solo mode, `onboarding_interviewer`'s first user turn is prefixed with a short hint (`interview.solo_hint(role_kind)`, one static constant per role kind from the onboarding role step): the draft names only the principal, as principal, and no departments; an owner is asked about the offer, customers, pricing, cash and runway, an in-house executive about their organisation (drafted as the company — their employer), function, remit, manager, team size as context and what they are measured on, an independent about their practice, clients and pricing, and with no role given the role-neutral `SOLO_HINT` asks which they are. The hint rides in the user turn and the tool array is unchanged, so the cached system prompt and tool prefix are the same in both modes.\n\n**Voice.** The Executive answers in one voice persona, substituted into block 0 at `{VOICE_PERSONA}` (`personas/loader.get_active_body`) from the Executive override's `voice_persona_slug`. Three general voices ship in `personas/builtin/`: **Direct** (`default`, the slug used when none is set: answer first, brief and decisive), **Supportive** (a warm coach who explains the why and checks how the plan sits) and **Analytical** (numbers first, options and trade-offs, then one recommendation). They are offered by the picker in Settings → Executive (setup doesn't ask), which saves through `PATCH /agents/executive`; a built-in voice's frontmatter carries the picker's one-line `description` and `sample` reply. The named-executive voices are marked `legacy: true`: the picker hides them, but they still load, so an install that picked one keeps it, and the Agent Council lists them under \"Legacy voices\". Voice files hold disposition only; the response-length ladder lives in `prompts/executive_persona.py`, and a test keeps numeric length rules out of every voice file. An eval scenario may set `voice_persona` to answer in a voice; it rides on `Session.voice_persona_slug`, which wins over the override for that session only.\n\nOne more structured-output utility agent sits beside `triage`: `openexecutive.agents.alert_review.AlertReviewAgent` (routing model, one forced `emit_alert_reviews` tool call, never raises). The alert review job hands it a batch of open alerts with evidence and gets back, per alert, a verdict (relevant / changed / resolved / stale), a confidence, the evidence it relied on, a one-line note, why-now, an optional deadline, and a recommended move (none / route / nudge / escalate / draft / suggest_workflow / merge / close) with its parameters. Like `triage`, it never mutates state — `alerts.review.apply_verdict` executes the verdicts deterministically within authority, so the model's job is judgment, not action.\n\n**Goals without departments.** The goal tools (`orchestrator.department_tools`) no longer assume a department: `list_department_goals` with no `department_slug` lists every goal with its owner and department, `update_department_goal` names a goal by `goal_id` alone, and `create_goal` takes an optional `area` and `owner_person_id` (a team member). The org block renders goals with no department with their `goal_id`, and their action chips link to `/goals` (see Org).\n\n**Solo mode.** When the workspace (or the turn's session) is solo — one person using Open Executive for themselves — the Executive works for that one person, the principal, whatever their role: they may run their own business, lead a function inside a larger organisation, or work independently, and the persona reads which from the context instead of assuming it — the role lines under \"Your Principal\" in the org block, when the principal gave their role (workspace settings: kind, title, reports to, remit, measured on). Its persona swaps the team sections (\"You Are a Member of This Executive Team\", \"Choosing Who to Tell\", \"Departments & Org Coordination\") for \"You Work for One Person\", \"The People in Their World\", \"Who Hears From You\" and \"Goals and Areas\". Solo means only the principal uses Open Executive — not that they have no people: their manager, direct reports, peers, clients, investors and board are contacts. The Executive helps the principal work with them (a 1:1 or review prep, an update or budget case for their manager or board, a negotiation) but does not coordinate them, contacts them only when the principal asks and only if they are a contact, and never offers to loop the principal in. Every specialist stays relevant — HR for the principal's own reports, board and investor communications for reporting upward. Goals are grouped by *area* (the tools' `department_slug` is the area's slug). The chat loop withholds `send_department_message`, `send_company_broadcast` and `set_department_head` from its tool list, and a call to one of them anyway gets a JSON error tool_result instead of running. The mode is resolved once per turn and pinned on the session, so the persona, the org block, the tools and every tool handler agree — a switch made mid-turn applies from the next turn. Team mode is unchanged.\n\nEvery specialist *advises* — each returns domain analysis the Executive synthesizes into one voice. None of them act on the company's behalf; the side-effecting tools in the main loop (schedule, people, departments, broadcast, watchlist, workflows) belong to the Executive, not to a specialist.\n\n`recall_history` (Always in the loop, `orchestrator.history_tools`) reads the speaker's own notes, on the same terms as `ghostwrite_email` below: its own registry, offered only on a verified turn in a conversation private to them with their switch on; its `due` flag lists only dated notes, soonest first (see Memory). The ghostwriter below also gets the writer's notes from mail every recipient was on, as a consistency check (`<writer_noted>`).\n\nThe morning reflection, which has no question to look up and no specialist to consult, gets one more read-only tool of its own: `search_knowledge` (`orchestrator.knowledge_tools`), the same knowledge lookup a chat turn makes, for what the company's documents say about a customer, contract, price or policy before it acts.\n\nA fifth, `remind_me` (`orchestrator.reminder_tools`), sets a plain-text reminder sent to the speaker alone when it is due; it opens no mailbox, so it never locks a turn. A sixth, `propose_actions` (`orchestrator.action_card_tools`), does nothing itself: it leaves the speaker a card of up to five actions (a meeting, a message, a new contact) that happen only when they tap Approve, so it stays on after a turn read mail; it is where what an email itself asks for goes, since that is never done on the email's say-so. Four read-only tools beside it read the speaker's own mailbox on the same terms — `search_my_email`, `read_my_email`, `read_my_email_attachment`, `my_email_awaiting_reply` (`orchestrator.mail_read_tools`) — and `my_calendar` their own calendar while Read my calendar is on, capped per turn, other people's messages and attached files returned as untrusted blocks (see Act as me). Beside them, `save_my_email_attachment` copies one attached file, when the speaker asked in their own words, into a fixed folder of the Executive's own Drive or OneDrive and shares it with them alone, every call built in codeOne tool writes under someone else's name, and it is fenced in code rather than the prompt. `ghostwrite_email` (**Act as me**, `orchestrator.delegation_tools`) writes an email as the person the Executive is speaking with and saves it as a draft in their own mailbox, Gmail or Outlook (`delegation.gmail.gmail_for` picks the client by the person's credential); chat never sends it (the person sends it themselves, or taps Send on its card). It lives in its own registry, never `_ALL_SKILL_TOOLS`, so reflection, research and workflow toolkits can never carry it. The chat loop adds it only on a turn `delegation.settings.pin_turn_delegation` offered it to — the principal, or a team member once the owner lets them, with it on, verified, in a conversation private to them (the web chat, a Slack or Discord DM, a verified private Telegram chat), on an attended turn — and builds that turn's handler map from both registries, so on any other turn a call to it is an unknown tool. From the round that calls it until the turn ends, the dispatch guard refuses only what opens an outside address, runs a script or workflow, or posts to everyone (`delegation.lockdown`), without changing the offered list; messages and invites to the roster, follow-ups, facts in the person's own words and a contact they name still run, each fenced in its own path. A later turn in the conversation stays private to its owner but holds nothing: the person asking again, having read the reply, is their own request. A contact add the lockdown refuses becomes an approval card in the chat instead (`propose_actions`). Recipients are chosen in code, never by the model (a draft's new email may go to any address, named back to check when it is neither on the roster nor typed), and the text comes from the **ghostwriter** (`delegation.ghostwriter`), a tool-less call with a constant prompt plus the person's \"How I write\" profile. The ghostwriter is deliberately not a Council agent: the Council's edit routes have no principal check. See **Act as Me (Delegation)**. A second tool joins the same way: `recall_history` (`orchestrator.history_tools`, Always in the loop) reads the speaker's own notes of what they said, by email or in chat. It has its own registry and per-turn handler map too, and the loop adds it only when `history_tools.recall_person` finds the speaker verified, in a conversation private to them, on an interactive turn that isn't private to the principal, with their notes switch on. It doesn't need Act as me. After every chat turn the Executive also hands the speaker's own words to `history_chat.schedule_chat_notes`, which takes notes in the background when their switch is on and the surface verified them (see **Memory**). A meeting `create_calendar_event` proposes works the same way when the person chatting is the one who approves it: its chip carries the `decision_id` and the chat shows the meeting card (Approve, Edit, Dismiss) under the reply, while the briefing keeps it too. A tool result held for someone else's approval (`waiting_for_approval`, the department gate) gets a *Waiting for approval* chip, never a done one.\n\nThe Executive can also launch *any* workflow conversationally: `orchestrator.workflow_run_tools` adds `list_workflows` (the launchable catalog with each workflow's inputs) and `run_workflow` (runs a built-in or custom workflow to completion and returns its artifact, checkpointing approval gates as `awaiting_human`). `run_workflow` runs the weekly review, and in solo mode the morning brief, only for the principal on a verified surface (`Workflow.principal_only_modes`, `create_goal`'s rule); anyone else gets an audited refusal that does not name the rule. A run whose `result` event reports `private_to_principal` (the morning brief on the principal's own verified turn, when it read their private rows) still returns its artifact into that turn, but the stored run keeps `persistence.PRIVATE_RUN_ARTIFACT` instead of the text, since run history is readable by everyone signed in. This makes the `/jobs` catalog reachable from chat — see the Workflows section.\n\nDeliverables are first-class too. `orchestrator.artifact_tools` gives the Executive three tools: `draft_artifact` publishes a memo, web page (`html`), Word document (`docx`), spreadsheet (`xlsx`, from structured `sheets`) or a `link` to something it created in one of the principal's connected apps through the generic MCP gateway (`search_tools` → `call_tool`, any server — no vendor is named), and returns the artifact's id and link; `list_artifacts` and `get_artifact` let it find and reread its own past work (drafts and workflow outputs alike, through the same `artifact_records` read model the API uses). Every artifact belongs to someone: a draft to the person whose turn published it (`artifact_records.current_viewer`; with nobody speaking, the scheduler or the CLI, the principal), and a run someone started by hand to them, while scheduled runs are the team's. The tools, the recall, `message_person`'s link and mail attachments only ever reach the speaker's own (a link is refused when the recipient couldn't open it), not even the principal sees a teammate's, and a speaker who isn't on the People list publishes nothing. A workflow's own session inside someone's turn (the research synthesis) is pinned to that turn's viewer (`Session.documents_viewer`). Revising is `draft_artifact(supersedes=<id>)`, which archives the earlier version. When the principal asks for a deliverable the tool is always used; the \"quiet is the right default\" bar applies only to unprompted drafts. The format enum is a hard-coded sorted tuple so the cached tool schema stays byte-stable. Delivery: `message_person(person_id, text, artifact_id?)` appends the artifact's title and a `UI_BASE_URL/artifacts/<id>` link (through the same outbound guard as any DM; the link needs Open Executive access), and a Gmail send/draft can attach the file itself by `artifact_id` — see Integrations.\n\n**Extensions** (`orchestrator.extensions`). An install can add chat tools and kinds of document without changing this code. `OPENEXECUTIVE_EXTENSIONS` names Python modules. The first turn, or the first Documents read, imports each one and calls its `register()`. A module that fails is logged and skipped. `register_tool` takes an `ExtraTool`: a definition, an async handler, an optional `tool_groups` group (new or existing) and an optional chip. It refuses a name the loop already knows (`executive.builtin_tool_names`). An extension tool is never in `_ALL_SKILL_TOOLS`, so reflection, research and the synthesis never see it. The chat loop adds it per turn, like the Act as me tools, but never on an unattended run, a turn private to the principal or to a speaker who isn't on the People list (mail from a stranger), and only where its own `offered(session)` says. A tool registered `known_caller=False` is the one exception to the People-list rule: it is asked on a turn with no verified caller too (a surface of the extension's own that names its user without proving who is speaking), so `register_tool` requires it to have an `offered` that checks who is asking; it is still never offered on an unattended, private, AI-app or replay turn. That predicate depends only on the kind of turn, so each kind of turn keeps one cached tool list. Act as me's lockdown refuses it once the turn has read the owner's mail, because it is unclassified and the lockdown fails closed, unless it was registered with `after_mail_read=True`: a tool that only reads the install's own data and carries nothing the model wrote anywhere outside. Its chip goes through `action_chips.summarize_action`, which applies the same no-chip rules for an error or refusal and keeps a link to a path inside the app. A `waiting_for_approval` result is the tool's own chip's to describe, never core's generic \"Waiting for approval\": the department gate never holds an extension tool. A tool may also name the chat's activity line while it runs (`activity_label`) and keep its dispatch audit row private (`private_row`: its input or result quotes the speaker's own data). `register_collection` adds a kind of document: a drafted artifact tagged `collection:<name>` carries that collection on its record. The Documents page gives each collection its own filter. The collection's `on_change` hears about archive, restore and delete before they happen; if it raises, the change does not happen. `draft_artifact(supersedes=…)` refuses a collection's document. The alert pipeline strips `collection:` tags, so only the extension's own code files a document in its collection. A module that fails partway through `register()` leaves nothing registered. A tool registered with `workflow=True` is also offered to workflow action steps as `oe__<name>` (see Workflows); `read_only=True` marks one that only reads. A step runs with nobody in the turn, so `offered` isn't asked: `extensions.in_workflow_step(name)` is True for that tool, during its call only, and the handler decides what an unattended call may do.\n\nDocuments come in the other way through `read_document` (`orchestrator.document_tools`): it reads a PDF, Word, Excel, CSV, Markdown or text file by `path` — one another tool saved, such as a Gmail attachment (`get_gmail_attachment_content`) or a Drive file (`get_drive_file_download_url`) — or a company document by `filename`, with an optional PDF `pages` range. A scanned PDF is converted (`knowledge.pdf_reader`: local OCR by default; with `PDF_PROVIDER_READING=true`, the deployment's own model through its provider's PDF support, falling back to local OCR) rather than returned empty. Reads are confined to the tool download folders — the same resolve-then-contain check as `oe__read_file` — plus the company docs folder, and it is withheld on turns private to the principal (`PRIVATE_TURN_WITHHELD_TOOLS`): those come from a contact's email, whose own attachments the poller has already read into the turn, and must not steer the Executive into company documents.\n\nThe Executive also backs the **Ask OE** panel's form-fill tier via `orchestrator.form_tools.propose_form_values`: when a chat turn arrives with a `<page_context>` block describing a form on the user's screen, the Executive can propose values for it. The proposal is delivered to the UI as a `form_patch` SSE event and placed into the form as highlighted suggestions — the tool never saves or submits anything; the user reviews and saves manually. The tool is **always** in the tool list (tool definitions sit at the top of the cached prompt prefix, so a conditionally-present tool would invalidate the prompt cache whenever panel and non-panel turns alternate); its description restricts usage to turns that carry a form descriptor. For questions about the product itself, the always-offered `read_user_guide` tool (`orchestrator/guide_tools.py`) returns any `guide/prebuilt/<id>.json` section; its description lists the section ids, fixed at import, so it never changes the cached tools prefix.\n\n**Take the lead** (`orchestrator.take_the_lead`) is how the Executive acts on its own, behind one gate. It has two switches, both off until someone turns them on. **As the Executive** is the principal's alone (`PUT /take-the-lead`). With it on, the unattended passes (reflection, research, the scheduler's proactive trigger) keep the acting tools solo mode otherwise withholds from them: booking, starting workflows, and messaging teammates rather than only the principal. `schedule_tools.unattended_toolkit` and the chat loop's unattended path then wrap every acting tool (`ACTING_TOOLS`, plus any MCP tool whose name sends, creates, updates, deletes or shares) in `gated_handlers` / `gated_call_tool`. The principal-only tools (`UNATTENDED_WITHHELD_TOOLS`) stay withheld, and every guard inside a handler still applies, including department approval levels (a booking in a department on Proposes is still a proposal) and the contact-egress rule. New mail or chat (any turn with an `origin_channel`) wakes the existing reflection early (`take_the_lead.wake`): one run waiting at most, five minutes after the first message and no more than every 30 minutes, as a one-off `take_the_lead_wake` scheduled action. Pause holds it like any scheduled work. **As you** is the Handle it for me dial's top step (delegation section).\n\n**The gate** is plain code (`take_the_lead.check`). Six kinds of action wait for a yes first, each with an Ask first switch, all on to start. The switch is the decision ledger's own mode for the class `take_the_lead_action:<kind>` (`propose` asks first, the default; `auto_execute` doesn't). The kinds are money, contracts and legal, hiring and people decisions, deleting/cancelling/sharing, someone new (an address on neither the roster nor the contacts) and big sends (a broadcast or department message, or more than five people). On top of these, the company (set by the principal) and each person (for their own As you) add structured rules: a person or address, a domain, words, or an amount at or above a figure. Added rules always hold, and an error reading them holds the action. The one thing the gate lets straight through is a direct message (`message_person` or a Slack, Telegram or Discord DM) whose only recipient is the principal (`_only_to_principal`): they are who it would wait for, so a card would only ask them to approve a message to themselves. Anything that reaches or acts on anyone else is checked, and a roster that can't be read is not the principal. A held action becomes a `take_the_lead_action` decision instance with a companion `DECISION_ALERT_SOURCE` card on Today, routed to the first teammate holding one of its People approval ranges (`take_the_lead.approval_ranges`, plain code, closest fit first: money to `spend_lt_2k`, `spend_lt_10k` or `spend_gt_10k` by its largest amount, with a higher range able to approve a lower one and no amount read as large, or to `customer_credit` for credit and payment terms; vendor contracts to `vendor_onboarding` then `legal_sign`, other contracts to `legal_sign`; hiring to `hiring_signoff`; anything else about the board or investors to `board_comms`; in Personal as in team mode), or else the principal through the authority gate's fallback (`departments.authority._route_proposal`). Deletes and shares always go to the principal. Only the principal and that approver see or resolve it (`DecisionClassSpec.principal_and_approver`). Approving it claims the row and carries out the exact call once (`take_the_lead.carry_out`: the unwrapped handler, or the gateway's `call_tool`, on an unattended session, with contact egress only when the principal approves). Declining drops it. Everything done or held is a row in `take_the_lead_log`. What it did and what was approved show in Recent activity (`today._build_activity`, kind `took_the_lead`), named but never quoted, since that feed is the team's. The held card's own `decision_resolved` row is left out because its payload quotes the message. `GET /today/activity` also adds the caller's own Handle it sends (kind `sent_as_you`, actor `you`), which are never in the shared feed.\n\n**In training.** The As the Executive switch has three positions: Off, **In training** and On (`take_the_lead.training`). In training, an action the gate would let through still waits, as a `training` hold for the principal alone, unless the principal has allowed that action with that person or thing (`take_the_lead.allowance`: \"Message Priya Nair\" covers a message to her on any channel; \"Book meetings with Priya Nair and Sam Lee\" meetings with exactly them; a workflow by id; a connected tool by name and the file it names, never one that reaches people such as a mail send or a share). The card carries Approve with a **Do this on its own next time** tick (`allow`), **Edit** (the message, a meeting's title, times or description, a task and its due date: `take_the_lead.apply_edits`) and Decline. Allowing is the principal's alone and needs a request the API can tie to them, as turning it on does; the six kinds and the added rules hold whatever is allowed. An edit approved with \"Do it like this next time\" is kept as an example (only the edited text, the principal's own words), and `take_the_lead.learned_note` hands the examples and the fact it's in training to the reflection and research passes in their user turn. After `SUGGEST_AFTER` (3) unchanged approvals of the same action in 30 days, Settings suggests allowing it. What it's learned is one list keyed by feature (`take_the_lead_allowed.feature`), so other features that learn from approvals can join it; each keeps its own switch. Act as me is the second (`act_as_me`, per person, `delegation.training`), with a training switch per setting (Replies, Follow-ups, Suggested actions, Drafts): its rows carry the person's id and the setting in their key, are listed and removed by that person alone, count toward their own cap, and never reach `learned_note`.\n\n**AI disclosure.** Nothing the Executive sends to people outside the web chat goes without saying it is an AI: plain code adds the line in the MCP gateway (mail) and at each chat send (Slack, Telegram, Discord, Google Chat replies, its DMs and room posts), never the prompt, so the persona and the cached system blocks are unchanged and a reply the model writes can't drop it. Replies the inbox watcher writes as a person carry their own line by that person's switches. See Integrations → **AI disclosure** and Act as me → **AI line on replies**.\n\n**App control from chat** (`orchestrator/app_control.py`, catalog in `orchestrator/app_control_catalog.py`). Three tools in the `app` tool group let the Executive do what the app's buttons do when the person asks (\"pin this\", \"delete that chat\", \"turn off Handle it for me\"): `find_in_app(kind, query)` lists what the speaker may change (their chats, topics, documents, workflow runs, settings; files, memory, departments, skills and workflows for the principal), `do_in_app(action, id, value)` does one action (`id: \"this\"` is the chat they are in), and `open_in_app(screen)` returns an in-app link. Each action calls the API route its button calls, as a function, with a request naming the turn's verified speaker (`correction_tools._speaker`; an email or unattended turn gets a refusal), so the route's own checks decide what this person may do. Actions whose route checks nothing about the caller (their page is the principal's) are `principal_only`. Deletes, shares, turning on Act as me, Handle it for me or Take the lead, removing a Take the lead rule, clearing every card and approving a workflow's waiting step never run in the turn: they become an `app_action` decision card, private to the speaker, shown under the reply; the tap runs the action with the tapper's own request through `POST /decisions/{id}/approve` (claimed once; a card older than a day is turned down). After a turn reads outside text or the speaker's mail, every action becomes a card. The tools are withheld from unattended and AI-app turns, and `do_in_app` from a shared chat someone else spoke in. Snoozing a card uses `ack_alert`'s trust rule: only ids the server showed or found this turn. Extensions add kinds, actions and screens with `register_kind` / `register_action` / `register_screen` at startup. The person's notes stay with `recall_history`; account, sign-in and billing stay on their screens.","mermaid": "flowchart TD\n User([User message]) --> Exec[\"Executive<br/>orchestrator\"]\n Exec -->|consult_specialist| Cso[\"cso: StrategyAgent\"]\n Exec -->|consult_specialist| Cfo[\"cfo: FinanceAgent\"]\n Exec -->|consult_specialist| Gc[\"gc: LegalAgent\"]\n Exec -->|consult_specialist| More[\"chro / coo / cmo / cpo / sales<br/>board_comms / triage\"]\n Mode{{\"Team or solo<br/>(persona + tools per turn)\"}} --> Exec\n Exec -->|\"ghostwrite_email (Act as me)\"| Gw[\"ghostwriter<br/>(tool-less, not a Council agent)\"]\n Cso --> Claude[\"Claude (Anthropic)\"]\n Cfo --> Claude\n Gc --> Claude\n More --> Claude\n Claude --> Exec\n class User entry\n class Exec compute\n class Mode compute\n class Cso compute\n class Cfo compute\n class Gc compute\n class More compute\n class Claude external\n class Gw compute",- "generated_at": "2026-10-21T00:00:00Z"+ "generated_at": "2026-10-21T16:00:00Z"}diff --git a/packages/core/openexecutive/architecture/prebuilt/api.json b/packages/core/openexecutive/architecture/prebuilt/api.jsonindex f9d9a763..fd7034c8 100644--- a/packages/core/openexecutive/architecture/prebuilt/api.json+++ b/packages/core/openexecutive/architecture/prebuilt/api.json@@ -1,7 +1,7 @@{"section_id": "api","title": "API Reference",- "markdown": "The FastAPI surface lives in `openexecutive.api.main`, which mounts one router per topic. A shared-secret middleware gates every request via an `x-api-key` header, except `/health` and the `/webhook/*` callbacks (which carry their own verification). A request it lets through is marked `request.state.shared_secret_ok`, which is what lets an MCP request turn on \"may act\" with `x-mcp-client-actions: on` (`mcp_server.may_act`; with no secret the header is ignored and `MCP_CLIENT_ACTIONS` decides). With local login (`OE_LOCAL_LOGIN=1`, which only `make dev` sets, and only while neither Google nor SSO sign-in is set up) a second middleware answers 403 to any request whose raw `Host` is not `localhost`, `127.0.0.1` or `[::1]` (except a webhook that verifies its caller: Google Chat and Slack always, Telegram only with a `TELEGRAM_WEBHOOK_SECRET` Telegram can send), and to any write whose `Sec-Fetch-Site` is present and not `same-origin`/`none`. The UI signs that owner in with no password and this API usually has no secret, so without it a web page in the owner's browser could drive the API as the principal — by DNS-rebinding its own hostname to 127.0.0.1, or by posting a plain form from another site or localhost port. CORS is restricted to the UI origins in `BACKEND_ALLOWED_ORIGINS`, plus the `make dev` UI on `localhost:3000` only when `OE_PUBLIC_DEPLOYMENT` is unset, so a public instance never answers a localhost page with credentials. The gate is off when `BACKEND_SHARED_SECRET` is unset — intended for local development only — so any internet-reachable deployment sets `OE_PUBLIC_DEPLOYMENT`, and `create_app()` then refuses to boot without the secret rather than serving an unauthenticated API.\n\n**Who is calling** is read in one place, `api/caller.py`, which answers with a `Caller` kind and email; a test fails on any other read of an `x-caller-*` header. By default (`open`) it is the `x-caller-email` the UI proxy stamps from the sign-in, trusted because whoever holds the shared secret is trusted as the proxy, and a request with none runs as the principal. **Signed callers** remove that trust. With `CALLER_ASSERTION_PUBLIC_KEYS` set, the proxy signs every request with an Ed25519 key only the UI holds, and a third middleware checks the `x-caller-assertion` header:\n\n- **What it covers.** It is good for one method, the exact path and query, at most 120 seconds, and one use. The API keeps used ids in memory.\n- **What it names.** A `user`, by email, or the `operator`: local login, which runs as the principal.\n- **No caller header at all** is a `service`, for scripts or MCP clients holding only the secret. It is never the principal, not even on an install with no owner yet, so it can't save a first setup.\n- **Anything else is a 401.** That includes an `x-caller-email` without an assertion (`caller_assertion_required`), and a bad, expired, reused or mismatched assertion (`caller_assertion_invalid`).\n\nPublic keys that can't be read stop the boot, rather than fall back to trusting the header. `scripts/make-caller-keys.py` makes the pair, and `scripts/mint-caller-assertion.py` signs one request for curl. The API holds only public keys. Setup's API-protection check warns while a shared secret is set without them.\n\nEndpoints grouped by router:\n\n| Router | Key endpoints | Purpose |\n|---|---|---|\n| `chat` | `POST /chat`, `POST /chat/stop`, `POST /chat/add`, `POST /chat/background`, `GET /chat/voice`, `POST /chat/transcribe` | SSE stream of a live Executive turn, with a `: keepalive` comment every `CHAT_STREAM_KEEPALIVE_S` while the turn is quiet. `POST /chat/background` starts a new chat that the server answers to the end with nobody connected (a phone approving a card, then closing the app): it answers `202` with the new `session_id` once the same caller, access and quota checks as `/chat` pass, the turn saves itself as a `/chat` turn does, and one caller may have three running at once (`429` beyond that). `POST /chat/add` hands the running turn a message the person typed while it works; the turn takes it at its next step and the stream names it in a `message_added` event, and a second `POST /chat` on the same conversation waits for the first to finish. `GET /chat/voice` says whether the chat mic shows (`available`, `max_seconds`); `POST /chat/transcribe` takes one recording (multipart `file`) and answers `{text}` for the message box, nothing kept: `404` when speech to text isn't set up, `400` for a file that isn't audio, `413` over 25 MB, `502` when the speech server fails. The request body optionally carries `page_context` (route, page title, guide section id, and a form descriptor with current values) — sent only by the Ask OE side panel so the Executive can explain the current page or propose form values. When the Executive calls its `propose_form_values` tool, the stream emits a `form_patch` event (`{type, form_id, fields, rationale, iteration}`) that the panel applies into the on-screen form as highlighted suggestions; the user reviews and saves through the form's normal submit path. Each round of tool calls also emits an `activity` event (`{type, label, tool, iteration}`) naming what is running — progress only, forwarded verbatim and never persisted with the message, unlike an `action_taken` chip. A `session_id` in the body must be a chat the caller may use (the same ownership rule as `sessions`). If it isn't, the turn runs in a fresh chat instead, so nothing is written into someone else's session and the reply looks the same as for an unknown id. An unknown namespaced id (`slack:…`, `telegram:…` and so on) is ignored and a fresh chat is started, because those ids belong to the channel adapters and claiming one first would make the caller its owner. The client mints a `client_turn_id` and sends it with the turn; `POST /chat/stop` takes that id and halts the turn mid-flight, so the button is live from the moment Send is pressed rather than only once the first SSE byte arrives. A stopped turn emits a `stopped` event and then the usual `done` over the same still-open stream — never an `error` — and whatever streamed so far is persisted and flagged. The endpoint returns 404 both for an unknown id and for another caller's turn, so it cannot be used to probe which turns are live. An optional `memory_text` (1–2000 chars) is what peer memory records as the caller's words instead of `message` — the briefing handoffs send it because their seed quotes the Executive's own card; `POST /chat/upload` derives it from the typed text and filenames. Since it never reaches the transcript, the `chat_turn` audit row keeps it beside `message`. After the reply — a stopped or timed-out one too, before the `stopped` / `error` / `done` frames — a turn that looked at anything sends one `sources` event, `{type, session_id, sources: [{kind, title, url}], unavailable: [area]}`: `kind` is `company`, `knowledge`, `notion`, `research`, `document` or `web`, `url` is only ever an http(s) link or, for an earlier document, `/artifacts/<id>`, at most 12 sources (6 from the web, pages the reply cites first), and `unavailable` names parts of the analysis (\"finance\") whose specialist failed or said nothing every time it was asked. It is persisted with the reply as `chat_messages.sources`. When the turn held steps for an outside read that still wait on the speaker's go (`outside_content.pending_held`), it then sends one `held_steps` event, `{type, session_id, steps: [plain words]}` (not after a stop or a timeout), which the UI shows as Approve / Not now under the reply; Approve sends \"Go ahead\". `GET /sessions/{id}/messages` puts the same `held_steps` on the last reply while they still wait (never for a shared chat's member). A request that carries `x-caller-email` marks its turn signed in (`Session.web_caller_signed_in`), which Act as me needs on a deployment without local login. |\n| `sessions` | `GET /sessions`, `GET /sessions/shared-with-me`, `GET /sessions/topics`, `PATCH\\|DELETE /sessions/topics/{topic_id}`, `POST /sessions/topics/{topic_id}/merge`, `POST /sessions/topic-areas/rename`, `POST /sessions/topic-areas/ungroup`, `GET\\|PATCH /sessions/{id}`, `GET /sessions/{id}/messages`, `DELETE /sessions/{id}`, `POST /sessions/{id}/messages/{message_id}/feedback`, `GET /sessions/{id}/followup`, `GET\\|PUT /sessions/{id}/shares`, `DELETE /sessions/{id}/shares/me`, `GET /sessions/{id}/people`, `GET /sessions/{id}/live`, `POST /sessions/{id}/typing`, `GET /sessions/{id}/export`, `POST /sessions/{id}/rewind` | List, read and delete saved chat sessions (assistant messages carry their row `id` and any `feedback`). The list is scoped to the caller, and each chat carries `preview`: its last message as one plain line, Markdown marks stripped and cut at a word near 120 characters, with `You:` before the person's own words. With `?q=` it is a search: the caller's own chats whose title or message text contains the words (case-insensitive, `LIKE` with wildcards escaped), each with `match_count` (messages that contain them; 0 for a title-only match) and `snippet` (the newest matching message, around the match, prefixed `You:` or `Executive:`). Scoped exactly like the list, so nobody else's conversation can match. The Chats page searches this way and opens a result with `?find=`, so the chat starts on that message and steps through the others. Each of the owner's chats also carries `topic_id`, the topic it is filed under (`memory.chat_topics`). `GET /sessions/topics` lists the caller's topics (`id`, `name`, `chat_count`, `kept`, `pinned_at`, `area`, `updated_at`) and, when some of their chats have no topic yet or some topics no area, starts filing them in the background from their titles and answers `organizing: true` until done. `PATCH /sessions/topics/{topic_id}` renames (a name another of their topics has merges the two) or pins one, `POST …/merge` moves its chats into another (`into_id`), and `DELETE` ungroups it, leaving its chats under no topic; another person's topic is a 404. `POST /sessions/topic-areas/rename` (`area`, `name`) renames one of the caller's areas and `…/ungroup` (`area`) drops it, its topics standing alone; an area they don't have is a 404. `PATCH /sessions/{id}` also takes `topic_id` (null for none) or `topic_name` (made if new) to file the chat, which then is never refiled on its own. A chat-app conversation started with /new is left off the list until its first turn (`memory.conversations`). Every per-session route checks ownership, because session ids are guessable (`slack:dm:<user id>`, `telegram:<chat id>`): only the session's owner or the principal gets through, and a legacy row with no owner is the principal's alone. A teammate's own chat (a web chat, a DM, a private Telegram chat; not a channel or group conversation, `chat._shared_space`) is theirs alone, the principal included, as their mail with the Executive is. Anyone else gets the same 404 as for an unknown id, so the routes can't be used to find out which chats exist. The one other way in is having started the chat in this process (matched by Person id or verified email, so it survives the caller being added to the roster), which is how a caller the roster can't resolve keeps using their own chat. That record is in memory, so after a restart such an ownerless chat is refused for reads. The principal can continue someone else's ownerless chat without becoming its owner. Deleting a session also drops its in-memory copy, so a later turn naming that id starts fresh. The feedback route records 👍/👎 (`{\"feedback\": \"up\"|\"down\"|null}`) on one assistant reply; only the session's own caller or the principal may rate it. The rater is stored (`feedback_by_person_id`), and a 👎 on a reply to the rater's own message triggers a working-style pass for them (paced; see attunement). The chat stream's `done` event carries `message_id` for the persisted reply. The followup route returns `{\"suggestion\": string|null}` — one suggested next user message, written by the utility-fast model from the last few turns, that the chat composer shows as its greyed placeholder (Tab / → or the inline chip fills it in without sending). It is null unless the session ends on a persisted assistant reply, and null (uncached) when the model fails. Same ownership check as feedback; successful suggestions are cached in-process per (session, reply id). Messages from `GET /sessions/{id}/messages` carry each assistant reply's saved `sources` (the same shape as the chat stream's `sources` event, without `type`), so a reloaded chat shows them again. The owner (not the principal on someone else's chat; the principal for an ownerless one) also has the chat controls. `PATCH` takes `{title?, pinned?}`: a name they give is kept (`sessions.title_locked`, so the automatic title never replaces it) and a pinned chat lists first, newest pin first, with `pinned_at`. `PUT /sessions/{id}/shares {person_ids}` adds exactly those team members to it (not contacts, the archived or the owner: 422), audited `session_shared`; a chat that read its owner's own mail (`mail_private`) is never shared (409) and stops being open to those it was shared with before, and a chat-app conversation is never shared (409). Those added get it from `GET /sessions/shared-with-me` (with `shared_with_you`, `shared_by`, `shared_at`) and **write in it with the owner**: `POST /chat` with its id runs their turn as themselves (`chat._shared_member` widens the turn's access, never its ownership), and they may read it, its messages (without the owner's ratings) and its export. They can't rate, rename, delete, share or redo it, and `DELETE /sessions/{id}/shares/me` takes them out of it (audited `session_shared`, with them under `removed`). In a chat several people are in, messages carry `sender_person_id` and `sender_name` (a message from before it was shared is the owner's), `GET /sessions/{id}/people` returns `{you, people: [{person_id, full_name, owner}]}` (owner first; empty for a chat of one), and redo is refused (409) when the last message is someone else's. `GET /sessions/{id}/live` is a server-sent event stream each open page holds (`api.chat_live`, one in-process broker): `ready` first (with the running turn so far, if any), then `turn_started` (who sent what), the reply's `chunk` / `thinking` / `activity` / `action_taken` / `phase`, `turn_ended` (the page reads the chat again), `typing` (from `POST /sessions/{id}/typing`, at most once per 2 s per person) and `people`. `removed`, `deleted` and `resync` (a page that fell 1024 events behind) end it; the access check runs again on every 20 s keepalive. At most 12 pages per person, 40 per chat and 2000 per process (429). Archiving someone takes them out of every chat shared with them, and their open streams close at the next keepalive. The page closes it while its tab is hidden. `GET /sessions/{id}/export` returns the chat as a Markdown file (`memory.chat_export`: each message under its writer's People name or `Executive`, the reply's sources listed under it). `POST /sessions/{id}/rewind {last_message_id}` takes back the last turn for Try again and Edit: under the conversation's turn lock it deletes the last reply and the messages it answered, but only while that reply is still the chat's last row, and returns those messages' words (`{message}`), which the page sends again as an ordinary `/chat` turn through the usual gates. 409 while a turn is answering, for a stale reply id, and for a chat-app conversation (its reply also lives in Slack, Telegram or Discord); audited `session_rewound`. |\n| `today` | `GET /today`, `GET /today/activity`, `GET /today/activity/daily`, `GET /today/brief-delivery`, `GET /today/value`, `GET /today/top-three`, `GET /today/weekly-review`, `POST /today/seen` | Morning-brief aggregation (in the install's first week, for the principal, `first_week_card` `{needs_goal, findings: [{title, summary}], brief_tomorrow}` or null), recent-action rail, Pulse heartbeat heatmap (per-day activity counts), and the owner's month tally for Pulse's Done for you (`GET /today/value?period=month|week&offset=N`: counted lines beside the period before plus the finished reports, owner only, 403 otherwise; see Today). `GET /today` carries the viewer's cached \"What's going on\" `narrative` with `narrative_generated_at` (when it was written) and `narrative_stale` (a fresher one is being written — the page re-polls; see Today), the same text split for Home's Right now card (`narrative_lead`, `narrative_points` `[{text, since}]`, `narrative_move`), `narrative_seen_at` (this reader's last look, which `POST /today/seen` records and returns as `{seen_at}`) and, for the principal, `morning_brief` `{at, text}` and `calendar_next` `[{title, start, all_day}]`. A proposal backed by a roster request carries `roster_request` (the new sender's server-derived details and held-message previews; the principal's only) and is answered through the `people` routes below. `brief-delivery` returns the latest morning brief or end-of-day digest that didn't reach the owner, as `{brief, at, problem, fix, readable}` in plain language (`readable` is false when it couldn't be written, so there is nothing to read), or null — null while briefs go out, once the cause is fixed, while the scheduler is off, and for anyone but the owner (`chat._caller_is_principal_or_unclaimed`); it carries no address or channel id. `top-three` and `weekly-review` feed the solo Briefing's own cards and answer only in solo mode and only to the principal (`workflows._caller_is_the_principal`: a principal match or no caller header, never an unclaimed roster); anyone else, and team mode, get null. `top-three` returns `{items: [{key, kind, text, why, slot}]}` — the morning brief's top three from the same code (`build_top_three(principal_due_soon())`), `slot` a free block like `10:00–11:00`, `\"\"` when none is left, null when no calendar was read — and is kept off `GET /today` because its calendar read may take up to 4 s. `weekly-review` returns the latest completed `weekly_review` run as `{run_id, completed_at, period, top_three, excerpt}` (next week's top three as plain text; `excerpt` when that list is empty), or null when none has completed. |\n| `people` | `GET\\|POST\\|PATCH /people`, `POST /people/{id}/archive`, `GET /people/requests`, `GET /people/requests/{id}`, `POST /people/requests/{id}/approve`, `POST /people/requests/{id}/decline`, `GET\\|POST /people/{id}/open-loops` (a `POST` tells the owner once and returns `notified`: `sent` with the channel, `held`, `not_reached` or `skipped`; see Attunement), `POST /open-loops/{loop_id}/close`, `GET /people/{id}/outreach`, `GET\\|PUT\\|DELETE /people/{id}/attunement` | Roster CRUD and scope lookups. Every `Person` carries `kind` (`team` / `contact`). `GET /people` lists team members only unless `?include_contacts=true` (the People page asks for both; the department-head and workflow pickers keep the default); `POST` / `PATCH` accept `kind`, and a principal contact is a 422 either way. They also take `email_aliases` (other addresses the person writes from, at most 10; on `PATCH` the full list replaces the current one, omitted leaves it): an alias matches their mail and may be emailed but never signs in; a malformed one is a 422, and an alias — or a new primary `email` — that is already someone else's address or alias is a 409. Every `POST` / `PATCH` then closes any pending roster request whose sender now matches someone and replays its held messages. Adding, editing and archiving (`POST`, `PATCH`, `/archive`) are the principal's only, or anyone's while no one is principal yet so a first setup can add its owner (`chat._caller_is_principal_or_unclaimed`, the check behind `PUT /workspace`; a request with no `x-caller-email`, such as the CLI or local login, counts as the principal). Anyone else, a person editing their own row included, gets 403 before the person lookup, so it can't probe which ids exist. Contacts are private to the principal (`caller_is_principal`: the signed-in email is the principal's, or no `x-caller-email`; stricter than the owner rule above, so while no one is principal nobody is): for anyone else `include_contacts` is ignored, a contact's id is a 404 on `GET /people/{id}` and open-loops (exactly as for a missing id; `PATCH` and archive are already a 403 for them), and creating a contact or turning someone into one is a 403, even on a first setup with no principal yet. `GET /people/me` returns `{person_id, is_principal}` so the UI offers contacts to the principal only. `GET /people/by-scope/{token}` never returns a contact. Archiving a person closes their open loops. The open-loop routes list what a person owes (soonest due first) and close one (`reason`: done / not_needed / cancelled) — both the principal's or the owner's only (403 otherwise, via `people.store.is_principal_or_self`). `POST /people/{id}/open-loops` `{task, due_date?}` assigns that person a task (201 with the loop; `due_date` is a local date, due 17:00 in the user's timezone, clamped to 60 days, default 2 days): any active team member may assign anyone on the team, 403 for anyone else; 409 for an archived person, a contact (as seen by the principal), a person at the open-loop cap, a duplicate or open loops turned off; 422 for an empty task. Nobody is messaged at assignment — the owner is chased once it is due. `/outreach` returns, per kind of proactive outreach over 30 days, `sent` / `replied` / `acted` / `ignored` / `pending` (voided DMs are excluded) — same access rule, checked before the person lookup so it can't probe which ids exist. `/attunement` reads, replaces (`rules`: up to 4 strings, or omitted to only set `locked`) or resets the person's working-style rules — same access rule, 404 for an archived person; `PUT` rejects a rule that is not purely about style with 422. Archiving a person also drops their working style and frees their aliases. The **roster-request** routes (declared before `/people/{person_id}`) are the principal's alone (`caller_is_principal`; anyone else gets 404, as for a missing id): `GET /people/requests?status=pending|approved|linked|declined|expired|superseded|all` lists up to 100 (`id`, `channel`, `channel_ref`, sanitised `display_name`, `profile_email`, `on_company_domain`, `suggested_kind`, `suggested_person_id` / `_name`, `status`, `resolved_person_id` / `_kind`, `message_count`, timestamps, `ack_sent`, and `previews` of the held messages while pending — shown to the principal, never to a model); `approve` takes either `link_person_id` (+ `replace_channel_id`) or `full_name` + `kind` (no default: 422 without both) + optional `role`; `decline` takes no body. An answered request is a 409, a chat id or address already on someone else is a 409; the held messages are replayed (or dropped) in the background. |\n| `contacts` | `GET /contacts/changes?person_id=`, `PUT /contacts/keeper`, `POST /contacts/changes/seen`, `POST /contacts/changes/{id}/undo`, `POST /contacts/changes/{id}/accept`, `POST /contacts/changes/{id}/dismiss`, `GET /contacts/{person_id}/note` | The contact keeper (see Org), principal only (404 for anyone else). `GET /contacts/{person_id}/note` returns what is remembered about one contact or teammate (`note`, `topics`, `last_emailed_at`, `last_heard_at`, `last_met_at`, `next_meeting_at`, `next_meeting_title`, `emailed_count`, `met_count`, `note_written_at`, `updated_at`), 404 when there is nothing yet or they are not active. `GET` returns `{mode, status, last_run_at, changes}`: the switch (`on` / `ask` / `off`), how the last pass went (`ok`, `act_as_me_off`, `not_connected`, …) and the last 30 days of changes (each `id`, `person_id`, `action` add/update, `field`, `email`, `full_name`, `old_value`, `new_value`, `reason`, `evidence`, `status` applied/suggested/undone/dismissed, `seen`, `created_at`); with `person_id`, that person's every change. `PUT` sets the switch; `seen` is Got it; `undo` / `accept` / `dismiss` return the change, 409 when there is nothing to undo or decide. |\n| `chat-directory` | `GET /chat-directory/people`, `GET /chat-directory/channels`, `GET /chat-directory/slack/lookup?email=` | The Slack, Discord and Telegram people and rooms the bots can reach, by name, for the People and department pickers (`integrations/chat_directory.py`). Each answers `{slack, discord, telegram}`, every one `{configured, ok, options: [{id, name, detail, email}], truncated, members_available}`: `configured` false (and no options) when that bot has no token, `ok` false when the provider could not be read, `members_available` false when Discord refuses to list server members (the Server Members intent is off). `lookup` returns `{configured, match}`, the Slack person with that work email or null. Cached in-process for 60 s; tokens are never returned. The principal's alone, like changing the People list (`chat._caller_is_principal_or_unclaimed`); channels also open while no one is principal yet, people and lookup wait for one; a 403 otherwise, and the pickers then keep only the by-hand id field. |\n| `departments` | `GET\\|POST\\|PATCH\\|DELETE /departments`, `.../goals`, `GET\\|POST /goals`, `PATCH\\|DELETE /goals/{id}` | Department state and goals. A goal has an optional department and an optional `owner_person_id` (a non-archived team member, else 422). `GET /goals` lists every goal with `owner_id`/`owner_name` (its own owner, else its department's head, else the principal) and `department_title`. `POST /goals` takes a goal plus `area` (a name, ≤80): an existing department by slug or title, else a new one of that name (no specialist, as with `POST /departments`), else no department when blank. `PATCH /goals/{id}` edits any goal and may move it (`department_slug`, null = none; `owner_person_id`, null = follow the department) — only the fields sent change; `DELETE /goals/{id}` removes any goal. The department-scoped goal routes still reach only that department's goals. `DELETE /departments/{slug}` keeps its goals with no department, on their owners. `PATCH /departments/{slug}` also takes `watched_entities` (≤50 names, `[]` clears) — the entities the research watch policy may add watches for and route to this department. A `cadences` spec that does not parse is rejected with a 422 (empty means no cadence, i.e. the check-in is off); a changed `check_in` cancels the pending check-in and queues the next one on the new schedule. `GET` and `PATCH /departments/{slug}` return the department state plus `next_check_in_at` (ISO UTC of the pending check-in, `null` when it is off, the workspace is solo or none is queued); the list route does not carry it. A new `head_person_id` must be a team member: an unknown id and one of the principal's contacts get the same name-free 422 (`person_id N is not on the team.`), so the route — open to any signed-in user — never reveals that a contact exists; an unchanged head is not re-checked. `POST .../goals` requires only `key_result`: a missing or blank `period_value` is filled with the current period for `period_type` in the user's zone (the same label `create_goal` uses), and `target` defaults to empty. `PATCH .../goals/{id}` rejects a blank `key_result` or `period_value` with a 422; `target` may be cleared. |\n| `agents` | `GET /agents`, `GET /agents/models`, `GET /agents/models/options`, `GET /agents/presets`, `POST /agents/presets/{id}`, `PATCH /agents/{id}`, `POST /agents/{id}/test` | Specialist metadata and Council overrides. `/agents/models` is the flat model allowlist the PATCH validator uses; `/agents/models/options` returns the same ids in the same order, each tagged with `provider`, `provider_label`, `route` (`direct` / `openrouter` / `local`) and a display `label`, for the Council's Provider → Model picker. `AgentDetail` carries `instructions` (additional instructions, appended after `prompt`, which stays the built-in prompt or its replacement); `PATCH` and `/test` accept `instructions` (max 4,000 chars, blank clears it). `GET /agents/presets` lists the Fast, Balanced and Thorough quality presets (each with `available` and the resolved `model` for this install) plus `active` (null when the council is Custom), `base` and `custom_agents`; `POST /agents/presets/{id}` applies one by writing per-agent model and deep-reasoning overrides and returns the same shape (404 unknown preset, 409 when the install allows none of its models). Each `AgentMeta` carries `visibility` (`core` or `internal`) so the Council's simple view can list core agents first. |\n| `personas` | `GET\\|PUT\\|POST\\|DELETE /personas` | Executive persona definitions. |\n| `alerts` | `POST /alerts/{id}/approve`, `GET /alerts/approvals/{id}`, `POST /alerts/{id}/ack`, `POST /alerts/bulk-ack`, `POST /alerts/{id}/snooze`, `POST /alerts/{id}/reopen`, `POST /alerts/review` | Groom the briefing queue. `approve` (body `{edited_body?}`, 202) carries a proposal card out on the server and records who approved it and when (`alerts.approvals`, see Today): it queues the scheduled action a department-gated card held, or runs one Executive turn as the approver in a new chat (`session_id`). 409 for a card already approved or closed, a decision-backed card (approved at `/decisions`), a roster request or a drafted artifact; 403 unless the caller is the card's routed person or the principal (an unrouted card is the principal's); 429 when the caller already has three background turns running. `GET /alerts/approvals/{id}` reads one back (`running`, then `done` / `failed` with `outcome`) for its approver, the card's person or the principal (404 for anyone else). `snooze` (`until`, an ISO time up to 30 days ahead; `null` clears it, the Undo) hides an `unread` card until then by setting `snoozed_until` (409 on a closed card or a roster request); the TTL then counts from the snooze, so a card snoozed past its TTL still comes back. `ack` (`read`/`ack`/`dismissed`, optional `mute_topic`) also teaches a watch-sourced alert's watch (`trust_score`); `bulk-ack` takes explicit `alert_ids` and/or `older_than_days` + `category` (exempt sources skipped); `reopen` is the Undo for a `resolved` / `expired` / `dismissed` alert (never `ack` or an artifact / decision row; back to `unread` with every review stamp cleared and its TTL anchor refreshed, so it is live again even if older than its TTL); `review` runs the Executive's relevance review on demand (ignores the interval; still honours `ALERT_REVIEW_ENABLED`, the per-pass caps, the single-flight lock and per-alert idempotency) and returns the move counts. `bulk-ack` caps ids at 500, needs `older_than_days >= 1`, and applies watch feedback per row; `mute_topic` must be one of the alert's tags or a scoped `external:` / `department:` pattern. Alerts still surface only through the briefing (`/today` reads the live view) — there is no list, prefs, mutes, or live-stream endpoint. |\n| `artifacts` | `GET /artifacts`, `GET /artifacts/{id}`, `GET /artifacts/{id}/download?as=`, `POST /artifacts/{id}/archive`, `POST /artifacts/{id}/restore`, `DELETE /artifacts/{id}` | Browsable deliverables in any format — Markdown, HTML, Word, Excel, or a link into a connected app — with `format`, `downloads`, `external_url`, `supersedes_id` and, for an extension's kind of document, `collection` / `collection_label` on every summary (`orchestrator.extensions`; archive, restore and delete tell the collection first and answer 502 if it refuses). `download` renders the file (Word/Excel built at request time from the stored text; any Markdown artifact, workflow output included, can export `?as=docx`) and is always an attachment with `nosniff` and a sandbox CSP. Every route answers for the signed-in caller (`artifact_records.Viewer`): a drafted artifact is its owner's alone (`alerts.owner_person_id`, the person whose conversation published it; none = the principal's), a workflow output follows its run (`persistence.run_visible_to`), and anything the caller may not see is a 404 like a missing id, the principal included. `POST /alerts/{id}/ack` and `/reopen` answer 404 for someone else's drafted artifact the same way. |\n| `decisions` | `GET /decisions`, `POST /decisions/{id}/{approve,reject,cancel}`, `GET`/`PUT /decisions/classes/meeting_scheduling`, `GET /audit/reliability` | Approval-gated decision instances (e.g. calendar bookings). What approve/reject/cancel do is per class (`DECISION_CLASSES`, one `DecisionClassSpec` each; meeting booking books on approve and deletes on cancel; `delegation_reply`, an inbox reply card, is its approver's alone (`approver_only`) — hidden from everyone else, the principal included when it is a team member's — sends its Gmail draft on approve and, after a reject, deletes the draft unless it was edited; `delegation_invite`, a meeting invite card, is its approver's alone too, and answers the invite in their own calendar on approve (`edits.answer`) and nothing on reject; `department_action`, a team member's chat send or edit their department held (see Org), is the principal's and its approver's alone (`principal_and_approver`), makes that exact call on approve and audits `department_action_done` / `_failed` / `_declined`), and an instance of a class with no spec gets 409 instead of being carried out as another. Only the principal, or the person the decision went to for approval (`approver_person_id`), may approve, reject or cancel it. Anyone else gets 403, and a decision nobody was asked about, or one private to the principal, is the principal's alone. The resolver is recorded as `resolver_person_id`, and a reject that loses the race to another answer is a 409. `GET /audit/reliability` is the principal's alone (403 otherwise), because it counts private decisions too. A proposed booking is surfaced on the briefing as a companion alert; approve/reject/cancel here also clear that alert. An approval whose caller resolves to the principal runs the booking inside `grant_contact_egress()`, so a meeting the principal proposed with a contact can be created. Such a booking's payload is marked `private`: it is proposed to the principal, left out of anyone else's `GET /decisions`, and a 404 for them on `GET /decisions/{id}` and every action route. Reviewed in the briefing — there is no separate proposals page. The class routes read and set how meetings are handled: both return `{decision_class: \"meeting_scheduling\", mode: \"propose\" \\| \"auto_execute\"}` (default `propose`); the `PUT` body is `{mode}`, principal-only (403 otherwise, open before a principal exists — but then only `propose` can be set: `auto_execute` is a 409 until a non-archived principal is on the roster), 422 on any other mode, and a change writes a `decision_class_mode_changed` audit row. In solo mode the class mode decides a booking, and `auto_execute` applies only to a turn the principal started on a verified surface — any other turn gets a proposal. |\n| `delegation` | `GET`/`PUT /delegation`, `GET`/`PUT`/`DELETE /delegation/voice`, `POST /delegation/voice/learn`, `POST /delegation/voice/describe`, `POST /delegation/voice/signature`, `PUT /delegation/inbox`, `POST /delegation/inbox/check`, `GET /delegation/replies`, `PUT /delegation/replies/{id}/text`, `PUT /delegation/replies/{id}/recipients`, `POST /delegation/replies/{id}/answer`, `GET /delegation/actions`, `GET /delegation/invites`, `PUT /delegation/handle-it`, `PUT /delegation/disclosure`, `PUT /delegation/training`, `DELETE /delegation/learned/{id}`, `GET /delegation/handled`, `PUT /delegation/team` | Act as me — the **caller's own** delegation only; nobody sets it for anyone else. It answers the principal, and a team member once the owner lets them; anyone else gets 403 `not_available_yet`. `GET /delegation` returns `{enabled, gmail: {status, message, email, connect_command, outlook_connect_command, provider}, inbox: {enabled, status, message, watch_since, last_poll_at, checking}, handle_it: {enabled, mode, available, sent_today, lead, lead_available}, training: {replies, follow_ups, actions, drafts, learned}, disclosure: {enabled, handle_it, approved_sends, line}, team}`, where `team` is the owner's switch for the principal while `DELEGATION_TEAM_MEMBERS` allows it — `{enabled, members: [{person_id, name, enabled, inbox, drafts_30d, sent_30d}]}`, counts only — and `null` otherwise. `PUT /delegation/handle-it` takes `{enabled?, mode?}` (`mode` is `careful`, `balanced` or `bold`; anything else is a 422 `bad_mode`) from a caller the API knows is that person — 403 `not_yours`, 409 `caller_signing_required` with only the shared secret — and turning it on needs Act as me and Draft replies to my inbox on (409 `act_as_me_off` / `inbox_off`); `GET /delegation/handled` returns `{replies: [{decision_id, sent_at, to_name, to_email, subject, body, open_questions, gmail_link, source, summary}]}`, the caller's replies and follow-ups (`source` `follow_up`) sent on its own in the last 7 days, and the meeting invites it accepted on its own (`source` `invite`, `subject` the meeting's title, `summary` one line such as \"Accepted Dana's Pilot review, Wed 21 Oct, 14:30\"); a card on `GET /delegation/replies` carries `source` the same way, `can_allow` while Replies (Follow-ups, on a follow-up) is in training, and `learns_style` while Drafts is. `PUT /delegation/disclosure` takes `{handle_it?, approved_sends?}` (any other field is a 422), the caller's own switches for the AI line on replies sent as them (`disclosure.enabled` is the install's `AI_DISCLOSURE`); turning one on needs no signed sign-in, turning one off does (409 `caller_signing_required`), and a change is audited as `delegation_disclosure_changed`. `PUT /delegation/training` takes `{replies?, follow_ups?, actions?, drafts?, actions_on?}` (any other field is a 422; `actions_on` is Suggested actions On, and `actions` and `actions_on` set true clear each other) and returns the `GET /delegation` shape, whose `training` carries `actions_on` too; putting Replies or Follow-ups in, or Suggested actions out of training or On, needs no signed sign-in, anything else does (409 `caller_signing_required`), and putting Replies or Follow-ups in turns Take the lead as you off (`PUT /delegation/take-the-lead` on answers 409 `in_training` while they are). `training.learned` is `[{id, label, setting, example, uses, created_at}]`, what the caller allowed or kept in training (`setting` `replies`, `follow_ups`, `actions` or `drafts`; the kept example, or \"\"); `DELETE /delegation/learned/{id}` removes one of the caller's own (404 for anyone else's). `PUT /delegation/team` takes `{enabled}` from the principal alone (403 `principal_only`; 409 `not_available` when the install doesn't allow it). The Gmail `status` is `connected`, `not_configured`, `needs_reconnect`, `mismatch`, `no_email`, `shared_mailbox` or `error`, checked live with Google on each call. `PUT` takes `{enabled}` (any other field is a 422); turning it on needs the caller's own Gmail connected, otherwise a 409 whose code says why (`gmail_not_connected`, `gmail_needs_reconnect`, `gmail_mismatch`, `no_email`, `shared_mailbox`, `gmail_error`). The voice routes show, learn (from the caller's sent mail, one pass at a time — a 409 `in_progress`, `locked`, `too_soon`, `not_enough_mail`, `no_profile` or `changed` when it was locked or edited while learning, or a Gmail code), write it from the caller's own description (`POST /delegation/voice/describe` with `{description}`: one model call applies their words to the profile they have and returns it with a short `sample_reply`, saving nothing, so they review it and save it with `PUT`; a 409 `empty`, `too_long`, `in_progress`, `too_many` past 10 an hour or `no_profile`; no mailbox needed), edit and lock, and reset \"How I write\": a typed field that fails validation is a 422 `invalid_voice` listing each `rejected` field, and `clear_exemplars` / `clear_signature` drop those. `POST /delegation/voice/signature` reads the signature from the caller's Gmail settings again and changes nothing else (a locked profile stays locked), with the same Gmail codes as learning: a 409 when their mailbox isn't connected, it refuses the sign-in or the check itself fails (`gmail_error`), and a 502 `gmail_error` when reading the signature fails. On an Outlook mailbox it is a 409 `signature_unavailable`, since Outlook gives apps no signature to read, and the stored one is left alone. A caller `can_delegate` refuses — anyone but the principal, for now — gets 403 `not_available_yet`, which hides the Settings card. A request with no `x-caller-email` is a 403 `sign_in_required` unless the API runs under local login — a header-less call is not a sign-in (the caller is read through `api/caller.py`, as on every principal-only route). `PUT /delegation/inbox` takes `{enabled}` and turns **Draft replies to my inbox** on or off, returning the same shape as `GET /delegation`: on needs Act as me on (409 `act_as_me_off`) and the mailbox connected (the Gmail codes above, which keep their names for an Outlook mailbox too), and turning Act as me off turns it off too. `POST /delegation/inbox/check` starts a scan now in the background and answers 202 with the `inbox` state (409 `inbox_off` or `in_progress`). `GET /delegation/replies` returns `{cards: [...]}`, the caller's open reply cards from the database alone (no Gmail call): the decision id, the sender's name, address, relation and whether it was verified, the subject, when it arrived, what they wrote, a one-line `gist` (what they asked and what the draft says, written with the draft and brought up to date when an answer rewrites it; empty on an older card), the draft's recipients, subject and body, `open_questions`, `flags` and a Gmail link built from a fixed prefix. **Send** is `POST /decisions/{id}/approve`: it sends the card's draft exactly as it is in Gmail, after checking again that the caller is its owner — which on a server needs signed callers (409 `caller_signing_required` otherwise) — that both switches are on (`inbox_off`), no client slot is active (`client_slot`), the Gmail is connected (the Gmail codes), the draft is still there and unanswered (`draft_gone`, `you_replied`) and its recipients are allowed (`no_recipients`, `too_many_recipients`, `executive_recipient`, `not_from_you`). Recipients that changed, or a newer message in the thread, answer 409 `{code: \"confirm\", message, reasons, recipients}`; approving again with `edits: {recipients: <the list shown>, thread_moved_on: true}` is the second yes. Send with the *on its own next time* tick adds `edits.allow: true`: 422 `cant_allow` unless Replies (Follow-ups, on a follow-up card) is in training, 409 `learned_full` when their list is full, both before anything is sent; `edits.example: true` keeps a draft edited in the mailbox as how they write to that person while Drafts is in training. A draft edited in Gmail between those checks and the send is not sent (`draft_changed`). A double tap sends once (`already_handled`); Gmail refusing answers 429 `rate_limited` or 502 `gmail_error` with nothing sent, and a send Gmail didn't confirm answers 502 `send_unconfirmed` and is settled by the next inbox scan. Dismiss is `POST /decisions/{id}/reject`. `GET /delegation/invites` returns `{cards: [...]}`, the caller's own meeting invite cards (`delegation_invite`, Read my calendar on): `decision_id`, `status`, `created_at`, `from_name`, `from_email`, `relation`, `title`, `when` (in the workspace time zone), `start`, `end`, `all_day`, `clashes` (busy times only, `\"14:00–14:30\"`), `clash_checked`, `can_answer` (false for a sign-in that reads but can't answer the calendar; then `connect_command`), `can_allow` (Replies in training) and `waited_because`. **Accept** / **Decline** is `POST /decisions/{id}/approve` with `edits: {answer: \"accept\" | \"decline\"}` (accept is the default; `allow: true` with accept for the *Accept from them on its own next time* tick), on the same caller checks Send uses (403 `not_yours`, 409 `caller_signing_required`); 409 `calendar_write_needed` when the sign-in can't answer (the card stays), `gone` when the event left their calendar, `expired` once the meeting started, `already_handled`; 502 `calendar_error` leaves the card for another try. **Dismiss** is `POST /decisions/{id}/reject` and answers nothing. `GET /delegation/actions` returns `{cards: [...]}`, the caller's own approval cards (`delegation_actions`): the decision id, when it was made, a one-line `why`, and each action's `index`, `kind` (`invite`, `message`, `add_contact`), `summary` and `text` (a message as it will be sent, or a meeting's description). **Approve** is `POST /decisions/{id}/approve` with `edits: {only: [indexes]}` (and `allow: true` for the *Do this on its own next time* tick while Suggested actions is in training: 422 `cant_allow` otherwise or when nothing on it can be allowed, 409 `learned_full`; a card carries `can_allow`); it needs the card's own person signed in (403 `not_yours`, 409 `caller_signing_required`), Act as me on (409 `act_as_me_off`) and a card under a week old (409 `expired`), runs once (409 `already_handled`), and returns the decision with `final_payload_json.results`: per action `done`, `waiting` (sent to whoever approves meetings), `failed` with why, or `skipped`. Every change writes a private audit row. |\n| `history` | `GET /memories/history` (`?q=`), `PUT /memories/history/settings`, `PATCH`/`DELETE /memories/history/{id}`, `POST /memories/history/{id}/forget-conversation` | Always in the loop: the **caller's own** notes and switches only, resolved as for Act as me (a signed-in person on the roster, or the principal under local login; otherwise 403 `sign_in_required` / `not_on_roster`). `GET` returns `{notes, reply_notes, retention_days, effective_retention_days, company_retention_days, retention_choices, can_keep_notes, can_note_replies, can_set_company_retention, share_work_style, can_share_work_style}` (`can_keep_notes`: a team member on the People list may turn notes on; `can_note_replies`: they can use Act as me, so email replies are noted too; `share_work_style`: their own \"Share my work style with the team\" switch, `can_share_work_style`: a team member in a team workspace). `PUT` turns notes on only for someone who may keep them, and `share_work_style` on only for someone who may share it (403 `not_available` for either; turning either off always works), takes a person's retention only when shorter than the company's (422 `invalid_setting`), and the company retention only from the principal (403 `principal_only`). `PATCH` pins or corrects a note; `DELETE` forgets it; `forget-conversation` is \"Don't remember this\". A note that isn't the caller's is 404. Every change is audited privately to its person. |\n| `reflection` | `GET /memories/reflection` (`?person_id=`), `DELETE /memories/reflection/misses/{id}` | Learning: the turns the conversation review graded a miss (`memory.chat_review`), with the person's words, the reply, the tools run, what they wanted instead, the `kind` (`company` / `product`), the finer `cause` (null when the grade gave none) and the signals, plus `weeks` (`{week, graded, misses}` per ISO week, newest first) and `sees_everyone`. The caller is resolved as for `history` (403 `sign_in_required` / `not_on_roster`); each person sees their own, the principal those from chats `chat._session_access` lets them open (a teammate's own web chat or DM and a mail-private conversation stay their owner's; channels and groups are open), or one person's with `?person_id=` (a teammate naming someone else gets 403 `not_yours`); the weekly counts carry no words and are not filtered. The same response carries `lessons` (`memory/lessons.py`: live first, then retired, each with `shown`/`fine_after`/`miss_after` and `retired_reason`), filtered by the same rule, and `DELETE /memories/reflection/lessons/{id}` retires one (204; 404 for a lesson the caller may not see, already retired, or missing). `DELETE` forgets one miss (`is_principal_or_self`; 404 `not_found` otherwise) and writes a `chat_review` audit row. Each miss also carries its nightly replay (`memory/replay.py`): `replay_state` (`open` / `closed` / `evidence`), `replayed_at`, `closed_at`, `replay_notes`; and the response carries `replay` (`{misses, would_pass, evidence, last_run_at}`: of the caller's visible misses from the last 30 days, how many would go right today). |\n| `notifications` | `GET /notifications/settings`, `PUT /notifications/settings` | Settings → Notifications: the **caller's own** settings only, resolved as for the notes (a signed-in person on the People list, or the principal under local login; otherwise 403 `sign_in_required` / `not_on_roster`). `GET` returns `{quiet_hours: {enabled, start, end, days}, delivery, urgent_through, timezone, effective_timezone, held_count, gets_briefs, updated_at}` — `days` is `every_day` or `weekdays`, `delivery` `instant` or `digest`, `timezone` the zone they chose (`\"\"` for the workspace's), `held_count` how many messages wait for their quiet hours or brief (counted, never listed), `gets_briefs` whether they are the principal. `PUT` takes `{quiet_hours, delivery, urgent_through, timezone}` and replaces them (extra fields, equal start and end, a bad time or an unknown zone → 422); switching `digest` → `instant` makes what waited for the brief go now. Audited as `notification_settings_changed`. See Org. |\n| `audit` | `GET /audit/logs`, `GET /audit/logs/{id}`, `GET /audit/sessions/{id}`, `GET /audit/usage`, `POST /audit/log` | Durable audit trail of side effects. `/audit/sessions/{id}` adds a per-session token cost summary; `/audit/usage` aggregates token usage and OpenRouter cost across all sessions (totals, by day, by model, and by source — the part of the system that made the calls — including server-side web-search counts), plus `scripts`: how many `run_script` scripts ran (and worked, and in workflows), the tool calls they made, their time, and `turns_avoided` — an upper bound on the model turns those calls would otherwise have needed (each call past a script's first) — and `python_jobs`: how many Python jobs ran (and worked, ran a kept Custom tool, took a chat attachment), their time, CPU, largest peak memory and bytes in and out, and the tokens and charge of the turns that ran one. The Token usage page shows it as a Scripts section. Rows private to the principal (`private_to_principal`: a turn about their private mail, or a row naming one of their contacts on their own turn) are theirs alone: for any other caller, resolved as for `/alerts`, the usage totals leave them out. Once a principal is on the roster, anyone else reads only their own rows (a teammate's own turns, their Act as me work) in the list, by id and in a session, and the principal reads those as actions only — what was done, when and in whose conversation, with the text, details and `full` left out, and no text search over them; any other row's id is a 404, and a session shows only the rows the caller reads. |\n| `architecture` | `GET /architecture/sections`, `.../sections/{id}` | This documentation surface. |\n| `guide` | `GET /guide/sections`, `.../sections/{id}` | Plain-language user-guide content surface — mirrors `architecture` (static, version-controlled files; no LLM on the serving path). |\n| `health` | `GET /health`, `GET /health/honcho` | Liveness probes. Only `/health` is unauthenticated; `/health/honcho` sits behind the gate like everything else, because the exemption is an exact-path set. `/health` also carries `google_tools_missing`, the pinned Google tools startup priming could not find: `[]` once it has checked and they are all there, a list of names when a workspace-mcp rename or the configured services and tool tier left one out, and `null` for no answer — Google Workspace is not configured, or the probe has not finished or it failed. Only `[]` asserts they are present. It also carries `gateway`, the MCP gateway's state from its supervisor (`orchestrator.gateway_boot`): `state` is `off` (MCP not enabled), `no_servers` (enabled, the config names none), `starting` (an attempt is running), `up` or `down` (every attempt so far failed; `next_retry_at` is the next one), with `attempts`, `up_since` and `email_poller` (whether the poller that rides on the gateway runs); the failure's own text stays out, since the route is open to anyone, and goes to the log and the owner's notice on Home. A hosted instance whose gateway never came up used to answer `ok` here with no email going out or in. Only `[]` asserts the tools are present, so a probe that died cannot read as healthy; Setup status says which of the three `null` means. |\n| `version` | `GET /version` | The running version, the latest release on GitHub and `update_available`, for Settings → About. The GitHub answer is cached 6h, only a semver `tag_name` is used and the release link is built from it; `UPDATE_CHECK_ENABLED=false` skips the call and still returns the running version. No in-app updater: upgrading is host-side (`docs/deployment.md#upgrading`). |\n| `setup_status` | `GET /setup/status` | Settings → Setup status: one light per part of the install — AI model, company setup, owner, the Executive's address, API protection, Gmail, your own mailbox (Act as me), Slack, Discord, Telegram, Google Chat, the scheduler, the daily brief (when and where it goes) and Honcho — as `{checked_at, checks: [{id, label, state, summary, fix, link, last_activity}]}`, where `state` is `ok`, `warn`, `error` or `off` (optional and not set up), `fix` says what to do in plain language, `link` is an in-app page that helps (`/people`, `/onboard`) and `last_activity` is the channel's latest `integration_inbound` audit time. It tests live, with at most one free read-only call per service, each bounded to 5 s: Anthropic `models.list` (so an organisation key missing `ANTHROPIC_WORKSPACE_ID` shows up), Slack `auth.test`, Telegram `getMe` and `getWebhookInfo`, Discord `users/@me` only while the bot is still connecting, and — once the owner's own Gmail credential exists — Google's token refresh and Gmail `users/me/profile`. It also reads the listeners the lifespan leaves on `app.state` (Slack's socket, the Discord bot and its task, the MCP gateway), the scheduler heartbeat and the pause switch. When `google_workspace` is configured, the lifespan also starts `MCPGateway.prime_pinned_tools()` in the background (held on `app.state.mcp_prime_task`, cancelled at shutdown), so the pinned Google tools named in the prompt are callable without a search. With `DRIVE_SYNC_ENABLED` the lifespan also bootstraps the `drive_sync_scan` heartbeat beside Notion's (see Scheduler); the Drive folder sync has no HTTP route, only the `openexecutive sync-drive` / `purge-drive` CLI. With `CONFLUENCE_SYNC_ENABLED` it bootstraps the `confluence_sync_scan` heartbeat the same way; that sync is run by hand with `openexecutive sync-confluence` / `purge-confluence`. Sample values copied from `.env.example` are named as such without being sent anywhere. No token reaches a response or a log line: probes report a status code, a service's short error code or an exception's class name. Requests within 5 s of a run get that run's answer and concurrent requests share one run (kept per app on `app.state`), so the button can't become a stream of calls to those services. Gated like every other route; the UI adds a sign-in light of its own, from its own settings. |\n| `documents` | `POST\\|GET\\|DELETE /documents`, `GET /documents/{filename}`, `GET /documents/sources`, `POST /documents/sources/{drive\\|onedrive\\|notion}/sync`, `GET /documents/{drive\\|onedrive\\|notion}`, `GET /documents/{drive\\|onedrive\\|notion}/{id}` | Upload, list, view extracted text, and delete indexed company docs. `GET /documents` rows carry `source: \"upload\"` and the `domain` the upload was indexed under. The connected-source routes back the knowledge page's single Company documents list: `/documents/sources` gives each source's `{id, label, enabled, syncing, last_run, last_error, interval_minutes, file_count}`; `/documents/drive`, `/documents/onedrive` and `/documents/notion` list synced files from the sync state (`{enabled, last_run, last_error, files: [{id, name, url, modified_at, synced_at, indexed}]}`; a Drive row also carries `unreadable_reason`, why a file with `indexed: false` has no text — e.g. a scan OCR could not read — or null); `/documents/{drive\\|onedrive\\|notion}/{id}` returns the stored text (a OneDrive id is `<drive id>:<item id>`) for an id on record (the path comes from the state, never the caller). They are read-only — a synced file is removed where it lives. `POST …/sync` starts one tick of the same sync the scheduler runs as a background task (202); 409 when the source is not connected or a tick is already running (every sync holds a per-process run lock, so a manual run never overlaps the heartbeat; the route also keeps its own started task per source and re-checks just before starting one, so two simultaneous requests start one tick, not two), 429 with `Retry-After` within 60 s of the last successful run or of the end of the last tick in this process, failed ones included, so a failing or slow source can't be re-run back-to-back. The page's text view never loads images from synced content (a remote image would be a read beacon for whoever edits the shared folder). Upload is keyed on the filename, so re-uploading a document replaces its chunks rather than duplicating them, and `DELETE` removes the rows as well as the file. `domain` must be one of the nine specialist domains (including `sales`) or `general`; anything else is a 400 rather than a 200 that indexes where no specialist can retrieve. A scanned PDF is converted (`knowledge.pdf_reader`) before it is indexed and in the text view, as are scanned PDFs in `POST /chat/upload` and the client / onboarding intake uploads. The upload is streamed to disk 1 MB at a time and refused with a 413 once it passes 50 MB, and the file is parsed in a short-lived child process (see the RAG section), so a large document does not stay in the API process's memory. A Word or Excel file that cannot be parsed, or whose parse runs out of time, is a 422 with a readable reason. When every parser is busy, the response is a 503 with `Retry-After`. |\n| `executive` | `GET /executive/status`, `POST /executive/pause`, `POST /executive/resume`, `GET`/`PUT /executive/outside-content` | Operator pause switch for all autonomous work (scheduler, Gmail poller, workflow resumer); inbound chat and DMs stay live. Status returns `{paused, paused_at, paused_by, reason, held_actions, can_resume, name}` — `held_actions` counts due `pending` rows that fire on resume, and `name` is the Executive's name (`EXEC_DISPLAY_NAME`), which the web chat shows over its replies. `pause` takes an optional `reason` (≤200 chars) and is idempotent (re-pausing keeps the original start and reason); `paused_by` is the UI proxy's `x-caller-email`, else `api`. Behind the shared-secret middleware — not `SCHEDULED_ADMIN_TOKEN`, which the UI cannot send. Any caller may pause; `resume` is principal-only (`is_principal_or_self`, 403 otherwise; open while no principal is rostered; a header-less caller resolves to the principal as in `chat`), and status carries `can_resume` for the viewer so the UI hides Resume from everyone else. `GET /executive/outside-content` returns `{mode, can_change}` for the owner's **When I read outside content** setting (`ask`, `outsiders` or `trust`; `orchestrator.outside_mode`), and `PUT` with `{mode}` changes it for the principal only (403 otherwise), writing an `outside_content_mode_changed` audit row. |\n| `workspace` | `GET /workspace`, `PUT /workspace` | Solo/team mode, the user's time zone and the principal's role (`memory.workspace_settings`). Both return `{mode, timezone, effective_timezone, role_kind, role_title, reports_to, remit, measured_on, company_domains, company_domains_custom}` — `timezone` is what the user set (null when unset), `effective_timezone` the zone in effect (`timezone`, else `USER_TIMEZONE`, else UTC), and each role field is null until set. `PUT` is a partial update: only the fields sent change; `mode` is `solo` or `team` (not null), `timezone` an IANA zone, or null/blank to clear it; `role_kind` is `owner`, `in_house`, `independent` or `other`, and `role_title` / `reports_to` / `remit` / `measured_on` are free text (trimmed, at most 120 / 120 / 500 / 300 characters), each null or blank to clear it; an unknown zone, a bad mode or role kind, an over-long role field or an unknown field is a 422 and changes nothing. Switching to solo cancels every pending department check-in, switching to team re-schedules them, and a new zone re-times the pending briefs and reflection; a role change has no side effects (the next turn reads it). `company_domains` is the company's own email domains (a teammate matches by local part on them — see Org): `GET` returns the ones in effect (the stored list, or the domain of the principal's primary address, with `company_domains_custom` false) only to the same caller the role is shown to, `[]` for anyone else; `PUT` takes a list (at most 10 plain domain names, never a free-mail provider's — 422 otherwise, the value never quoted) or null / `[]` to go back to deriving them, and a change is audited with the new list. Each change writes a `workspace_settings_changed` audit row (a role change adds `details.role_fields_changed`, the names of the fields that changed — never their text). The role is the principal's: `GET` returns it only to a caller `PUT` would accept and reads it as nulls for anyone else. `PUT` is the principal's alone (`is_principal_or_self`, 403 otherwise; a header-less caller resolves to the principal as in `chat`) and open while no principal is rostered, so first-run setup can pick a mode. Both work before onboarding — unlike `company-profile`, nothing here needs a profile. |\n| `scheduled` | `GET\\|DELETE` scheduled actions (`?status=…&order=asc\\|desc`) | Pending and fired scheduled actions; `order` flips the `run_at` sort (asc=soonest-first for the upcoming queue, desc=most-recent-first for terminal-status history). `DELETE /scheduled/{id}` cancels a `pending` row (409 otherwise). Like pause it only stops work, so any signed-in user may — the UI proxy's `x-caller-email` is trusted when the request also carries the `BACKEND_SHARED_SECRET` `x-api-key` — as may a matching `X-Admin-Token` (`SCHEDULED_ADMIN_TOKEN`) or a loopback caller with no token set; otherwise 401 (token set) or 503. Audited as `scheduled_action_cancelled`. |\n| `watchlist` | `GET\\|POST\\|PATCH\\|DELETE /watchlist`, `POST /watchlist/describe`, `POST /watchlist/{slug}/approve`, `POST /watchlist/{slug}/decline` | External-monitoring watchlist + signals. `POST /watchlist/describe` `{text}` drafts a watch from a sentence (one forced tool call on the utility-fast model, `monitoring.describe`; the company profile's name, competitors, vendors and tickers ride in the user turn) and saves nothing: it returns the create fields (`slug`, `signal_type`, `target`, `display_label`, `trigger`, `cadence`, `mode`, `severity_floor`, `severity_ceiling`, `notes`) plus four plain lines (`watching`, `tell_you_when`, `how_often`, `how_loud`) written by code from those fields, after the create route's checks (registered kind, cadence, severity, the URL guard, the rss feed check, a free slug); 422 with a plain reason when it can't be a watch, 429 past 30 drafts an hour, 502/504 when the model fails. `GET /watchlist/{slug}` also returns `noise` `{recent_alerts, recent_dismissed, noisy}` (its last 8 alerts, noisy at 6 dismissed, `monitoring.noise`), and `PATCH` takes `quiet_noise_nudge: true` to stop that question for the alerts so far. `display_label` on create maps to the adapter's label key for every kind. Rows carry `origin` (`manual` \\| `executive` \\| `research` \\| `research_proposed`); `approve` turns a research suggestion (`research_proposed`, `dry_run`) into a live watch, `decline` `{reason: not_relevant \\| too_noisy \\| wrong_source}` removes it and records the decline (or, for `too_noisy`, keeps it live at a high floor), both 409 unless the row is a pending suggestion; `DELETE ?reason=` on a research-origin row records the decline too. |\n| `workflows` | `GET /workflows`, `POST /workflows/{name}/runs`, `GET\\|POST\\|PUT\\|DELETE /workflows/custom`, `GET\\|DELETE /workflows/custom/{name}/targets`, `POST /workflows/runs/{run_id}/decision`, `POST /workflows/runs/{run_id}/rerun`, `GET /workflows/tools/search?q=`, `GET /workflows/tools/describe?names=` | Workflow catalog, run management, and user-created (dynamic) workflow CRUD. `rerun` queues a failed scheduled run (a department check-in, a saved workflow on a cadence) once more, due now (see Scheduler); the principal may, and for a saved workflow the person it delivers to; 409 for a run that did not fail, one started from `/jobs` (\"Start this one again from its page\") or one already queued again. Custom definitions come back with a server-managed `owner_person_id` (the creator — the web caller on `POST`, set on insert only; a client-sent value is ignored), who is asked to approve a workflow's first write to a new target. `GET /workflows/custom/{name}/targets` lists the targets its tool steps may write to without asking and `DELETE …/targets?value=` forgets one. `POST /workflows/{name}/runs` is a 403 (\"Only the principal can run this workflow.\"), audited, before any run row exists, when the workflow is one only the principal may start in the workspace's mode (`weekly_review` always, `morning_brief` in solo) and the caller is not the principal. A request with no `x-caller-email` counts as the principal; with no principal on the roster, no signed-in caller does. This governs starting a run only. A run started by hand (this endpoint, or `run_workflow` / research in chat) records its starter as `owner_person_id` and is theirs alone: `GET /workflows/runs` lists the team's (scheduled, no owner) runs plus the caller's own and the ones waiting on them, and `GET` / `DELETE /workflows/runs/{run_id}` and `POST /workflows/runs/{run_id}/decision` are a 404 for anyone else's, the principal included. A signed-in email that is on nobody's People entry cannot start one (403), and a run whose starter can't be read is refused (503); its steps read and draft documents as the starter only. `POST /workflows/runs/{run_id}/decision` with `{decision: \"approve\"|\"reject\"}` answers a run awaiting a yes/no sign-off from the web app — 403 unless the caller is the person it waits on or the principal, 409 if it isn't awaiting or needs a written reply — through the same `apply_resolution` a chat reply uses. Custom create/update also rejects (422) any action-step tool that does not resolve, and so does `POST /workflows/custom/{name}/activate` when turning a workflow **on** — that click is the approval for a tool workflow chat saved switched off (turning off is never checked). Turning on a workflow with action steps requires the body `{is_active: true, definition}` carrying the definition the user reviewed; if the stored one differs (ignoring `is_active` and timestamps) it returns 409, and the switch itself is a compare-and-set on the validated revision (`dynamic_store.activate_if_unchanged`, keyed on `updated_at`), so a save that lands while the tool check awaits the gateway — even from another process — is a 409 too, so only what was shown gets switched on. `GET /workflows/custom` lists inactive definitions too; `GET /workflows/{name}` 404s for them. The two `tools` endpoints return `{tools: [{name, description, read_only, source}]}` for the advanced editor's tool picker and the review card; `read_only` is true only when the name says it reads (a label, not a control). The run stream adds a `progress` event (`step_id`, `summary`) while an action step uses tools. Each catalog entry (`WorkflowMeta`) carries `is_custom`, `background` (true for the system-run built-ins the UI files under “System”) and `quick` (true for a Quick workflow, one per playbook nothing else follows, named `quick_<playbook>`). The run endpoint streams SSE; its `awaiting_human` event carries `delivery` (`self` \\| `sent` \\| `suppressed` \\| `alerted` \\| `failed`) alongside `person_id` / `question` / `awaiting_until`, because the outbound guard or an unreachable approver means the question may never have been asked — a client must not render “waiting on them” for those. A paused run ends the stream with neither `done` nor `error`, so that `awaiting_human` frame carries `terminal: true` to say so explicitly, plus `resumable` — whether the run continues by itself once answered or stops at the gate. `GET /workflows/runs/{run_id}` returns the run row with the resume payload replaced by a small `resume_progress` summary (`gate_step_id`, `completed_step_ids`, or `null`): the payload holds every completed step's full text, and the run-detail page polls this endpoint while a run is unfinished. |\n| `knowledge` | `/knowledge/builtin`, `/external`, `/search` | MBA knowledge base and external sources. `/search` is the introspection parallel of `retriever.retrieve` and mirrors its filters — including the `general` catch-all on the company collection — so the panel cannot report a specialist seeing nothing that it does retrieve in chat. It accepts `general` in `domain_filter` for the same reason. `/search` applies the same review gate as `retrieve()` — withheld (pending or rejected) built-in, external and failure rows are dropped, so the panel never advertises knowledge the Executive cannot use; it degrades to unfiltered rather than failing if review state is unreadable. `POST /knowledge/builtin` refuses a name already held by shipped content, and `POST|DELETE /knowledge/failures` now register and deregister for review — without that a user-authored failure case study had no review row, so no SME decision could block it. |\n| `skills` | `/skills`, `/skills/search`, `/skills/{name}/restore` | Skill registry (the UI's **Playbooks**) CRUD and search. `PUT` on a built-in saves a company customization that shadows it; `DELETE` returns `{name, outcome}` — `deleted`, `reverted` (customization removed, built-in back) or `hidden` (built-in hidden for this company); `restore` un-hides; `GET /skills?include_hidden=true` lists hidden built-ins. List and detail carry `used_by` — `[{name, title, is_custom}]` of the workflows that follow the playbook (switched-off custom ones included) (`WorkflowMeta.playbooks` is the other direction). |\n| `evals` | `/evals/scenarios`, `/evals/runs` | Eval scenarios and run results. `POST /evals/runs` is a 403 when a workflow scenario names a workflow that only the principal may start in the mode it runs in (the weekly review; the morning brief in solo) and the caller is not the principal — the Jobs page's rule for starting a run; stored results are read under the existing `/evals/runs` rules. |\n| `memories` | `/memories/{decisions,initiatives,advice}`, `GET /memories/facts`, `POST /memories/facts/{fact_id}/{retire,approve,decline}`, `GET /memories/facts/approval`, `PUT /memories/facts/approval/{person_id}`, `GET /memories/people`, `GET /memories/people/{person_id}/conclusions` | Episodic-memory record management; the standing facts the Pulse Corrections tab reads (`{facts: [...], can_retire, retirable_ids, can_review}`; each fact carries `recorded_by_role` / `recorded_by_name`; `?include_inactive=` default true, honoured for the principal only; anyone else gets the active rows plus their own `proposed` ones, with the provenance (`session_id`, `turn_id`, `recorded_by_person_id`, `retired_reason`, and `source_quote` except on their own facts) blanked, never retired, replaced or declined text; `?limit=` 1–500) and their retire (the principal any fact, a teammate only one they recorded that the principal has not approved — else 403; the audit row names the teammate; 404 for an unknown or `profile` row, 409 when no longer active; writes a private `fact_retired` audit row); approve / decline a teammate's `proposed` fact (principal only, else 403; 409 when it is not proposed; private `fact_reviewed` row); and the per-teammate \"needs my approval\" switch (on by default; `GET` lists `[{person_id, full_name, needs_approval}]` for every active team member but the principal, `true` for anyone not yet set; `PUT` takes `{needs_approval}`, 404 for anyone else; both principal only; private `fact_approval_changed` row); plus the read-only peer-memory overview the Pulse People tab reads (`{status: ok|disabled|error, people: [...], conclusion_total}`; `?recent=` 1–50, default 5) and one person's full conclusion list, paged newest first (`{status, person_id, items, page, size, total, has_more}`; `?page=` ≥1, `?size=` 1–100, default 50; 404 when the person has no peer). Peer memory is each person's own: the overview holds only the caller's entry (none for a caller off the roster; `conclusion_total` counts only theirs), the principal included, and `/conclusions` for anyone but the caller is a 404. |\n| `review` | `/review/items`, `/review/stats`, `/review/trusted-defaults`, `POST /review/curate`, `POST /review/bulk-approve` | Knowledge review queue and annotations. Shipped knowledge registers as a trusted default (`approved` with a NULL `reviewed_at`), so the queue starts empty; `curate` (`{domain, action: start|stop}`, domain required) opts one domain in or out, and `trusted-defaults` returns domain → unreviewed count. `bulk-approve` takes a non-empty `domain`, `item_ids` (max 500) or `all_pending: true`, and 400s with no selector; it returns `{approved_count, item_ids}`. |\n| `onboarding` | `/onboard/interview/understand`, `/onboard/interview/start`, `/onboard/interview/message`, `/onboard/interview/draft`, `/onboard/interview/revise`, `/onboard/interview/{id}` (GET resumes, DELETE starts over), `/onboard/interview/commit` | Conversational company setup. The `onboarding_interviewer` agent asks clarifying questions, then emits a draft profile + people + departments. **Nothing is persisted until `POST /onboard/interview/commit`** — the interview itself never writes. `start` is multipart (description + optional deck/one-pager uploads). `revise` changes one earlier answer (0 = the description) and runs the interview again from there. `commit` answers 403 when the draft would make a different person the owner and the caller is not the current owner (setup demotes the old principal, and any signed-in user can open it). It also takes an optional `owner_email` — the principal's sign-in email, which the review screen pre-fills from the signed-in user — checked before anything is written (422 if it is malformed, already another person's, or would replace the principal's existing email — setup only fills in a missing one; 403 if the caller is not the owner and it is not the address they signed in with, since the roster is the sign-in allow-list) and saved on the principal's People row, so the owner's chats resolve to them from the first turn. |\n| `workflow_designer` | `POST /workflows/designer/start`, `POST /workflows/designer/edit`, `POST /workflows/designer/message`, `POST /workflows/designer/draft`, `GET /workflows/designer/{session_id}` | The conversational New workflow wizard, and editing a saved workflow by conversation. `/edit` (`{name}`) opens a session on a saved custom workflow with no model call: the stored definition is the starting draft, and every response then carries `editing` (its name) and `original` (the stored definition as the session opened on it); drafts keep that name and its on/off state whatever the model writes (404 for an unknown workflow). The UI saves one with `POST /workflows/custom/{name}/save-edit` (`{session_id, definition}`): the server saves the session's own draft, never a client-sent body, and only when it equals the `definition` the user reviewed (a later message is a 409, so the click saves only what was shown); it re-validates it, keeps the stored on/off state, and writes only if the stored workflow is still the version the conversation opened on — otherwise 409, as it is for a session about another workflow or one with nothing drafted yet. It sits under `/workflows/custom` so anything that limits writes to custom workflows covers it too. The `workflow_designer` agent asks clarifying questions (`phase: \"question\"`, with `hint` and up to four suggested `options`), then returns `phase: \"draft\"` with `draft.definition` (exactly the body `POST /workflows/custom` takes), `summary`, and `assumptions`; every response carries the `transcript`. **Nothing is persisted here** — sessions are in memory, and the UI saves through `POST /workflows/custom`. Errors are fixed strings (422 over the 8,000-char message cap or the transcript cap, 404 for an expired session, 409 while another turn on the same session is still running, 502/504 for model failure or timeout). Mounted ahead of the `workflows` router. |\n| `python_jobs` | `GET /python-jobs/{job_id}/{name}` | A result file a Python job (`run_python_job`) wrote, the principal's alone (403 otherwise), served as a download (`Content-Disposition: attachment`, `Content-Security-Policy: sandbox`, `nosniff`, `Cache-Control: no-store`); 404 for a bad id or name or a file past its 7 days. |\n| `saved_tools` | `GET /saved-tools`, `GET\\|PUT /saved-tools/{name}`, `POST /saved-tools/{name}/rollback` | The scripts the Executive saved as tools (Settings → Advanced → Saved tools), the owner's alone (403 otherwise). The list gives each tool's name, description, on/off, current version, `kind` (`script` or `python`, a kept Python job) and the tools it calls, plus whether `SAVED_TOOLS_ENABLED` is on; the detail adds the script, every version and the last 20 runs. `PUT {enabled?, workflows?, version?}` turns one on or off (off also turns it off for workflows, so turning it back on never re-arms unattended runs), and turns the named `version` — the one the owner looked at, never whatever is current by then — on for workflows (or workflows off); a Python tool never turns on for workflows (409) — turning workflows on lets more run unattended, so it needs a request tied to the owner (signed sign-ins or local login, as Take the lead's switches do; 409 otherwise); `rollback {version}` makes that version the one that runs. Both are audited as `saved_tool_changed`. The Executive saves tools itself through `run_script`; there is no create route. |\n| `onboarding` (fallback) | `/onboard/start`, `/onboard/answer`, `/onboard/status/{id}` | The original 12-step form wizard, kept as a fallback (UI at `/onboard?mode=form`) and used by the `openexecutive onboard` CLI. Its final answer saves the profile and adds people with their emails and a principal, so once there is an owner only they can send it (403, and the session rolls back so the owner can still finish it). `/onboard/start` reads the workspace mode once for the session: solo skips the Team Size, Team Members and Fractional Executives steps. Each status response's `current_step` / `total_steps` count only the steps that session asks, and `optional` says whether the current step can be skipped. |\n| `company-profile` | `GET\\|PATCH /company-profile` | Structured company profile. |\n| `alerts` (private cards) | `POST /alerts/{id}/ack`, `/reopen`, `/alerts/bulk-ack` | An alert tagged `private:principal` (see Today) is a 404 on ack and reopen for anyone but the principal, and their bulk sweep skips it. A roster-request card (source `roster_request`) is a 409 on `ack` even for the principal — it is answered at `/people/requests/{id}`, since acking would leave the sender waiting with nothing to show it — and, as an exempt source, `bulk-ack` and `reopen` skip it. |\n| `auth` | `GET /auth/allowed-emails` | Non-archived **team members** with an email, offered to the UI's sign-in gate — a contact is never on it, so adding a client to reach them by email no longer lets them sign in. The gate takes the **union** of this roster and its own `ALLOWED_EMAILS` env list, so env entries always work (and are checked first, short-circuiting the fetch) and roster entries are additive — a roster change can never revoke a configured operator. With local login (`make dev` with Google sign-in not set up — the UI then listens on 127.0.0.1 only) the UI signs in without an email, sends no `x-caller-email`, and the backend resolves the caller to the principal, as it does for the CLI. |\n| `fixtures` | `/fixtures`, `/fixtures/{name}/load` | Demo-company fixture load/reset/snapshot. `POST /fixtures/reset` leaves no departments, like a new install (`departments_seeded` is always 0). |\n| `clients` | `/clients`, `/clients/generate`, `/clients/cockpit`, `/clients/save`, `/clients/{slug}/activate` | Client-company slots — multi-client mode for fractional use; create from current/blank/AI-generated intake draft (pasted notes and/or file attachments, multipart on `/clients/generate`), save-back switching, engagement metadata (PATCH), cross-client cockpit, delete. See the Client Companies section. |\n| webhooks | `POST /webhook/telegram`, `POST /webhook/google-chat`, `POST /webhook/slack` | Inbound integration callbacks (self-verified). `/webhook/slack` answers 503 unless `SLACK_SIGNING_SECRET` is set. |\n| MCP server | `POST\\|GET /mcp` | Open Executive exposed as an MCP server (Streamable-HTTP). Not a router — a mounted FastMCP sub-app. Same shared-secret gate (`x-api-key`); exposes company context as resources and the specialist council as tools, answered as the verified caller (`api.caller`) by the web app's rules. See the MCP Server section. | A card the Executive left for something asked in chat (`app_action`, `orchestrator/app_control.py`) is its person's alone too: its payload holds `action`, `id`, `value`, `title` and the `card` line shown; **Confirm** is `POST /decisions/{id}/approve`, which runs the action through the route its button calls with that request (a refusal comes back as that route's status and words, and the card is marked `failed`), and a card older than a day is turned down with 409.\n\n**What it does for you** (`api/routes/may_do.py`): `GET /what-it-may-do` is read-only. It returns `{paused, sections}`, three sections in order (`as_you`, `executive`, `departments`), each `{key, title, available, unavailable, notes, rows, footer, goes}`, and each row `{id, what, outcome, then, why, setting, items}`: `outcome` is `goes`, `waits`, `always_waits`, `drafts` or `off`, and `setting` names where to change it (`act_as_me`, `handle_it`, `take_the_lead`, `pause`, `department:<slug>`, or empty). `as_you` is the caller's own (`available` false, with why, when they can't have Act as me); `executive` is the principal's alone (`available` false for anyone else); `departments` is everyone's. Any team member on People may call it; a contact, someone not on People, or a request with no caller outside local login gets 403. See Delegation → What it does for you.\n\n**How far it goes as you** (`api/routes/delegation.py`): `PUT /delegation/reach` takes `{step}`, one of `off`, `when_asked`, `inbox`, `careful`, `balanced`, `bold`, `everything`, and answers the caller's delegation state as `GET /delegation` does. Going down turns the switches above the step off, highest first; going up runs each switch's own route in order (`PUT /delegation`, `PUT /delegation/inbox`, `PUT /delegation/handle-it`, `PUT /delegation/take-the-lead`), so the first refusal (409 `gmail_not_connected`, a signed sign-in missing, `in_training`, 403 for Everything when not the owner) stops it at a whole step. See Delegation → How far it goes as you.\n\n**Take the lead** (`api/routes/take_the_lead.py`): `GET /take-the-lead` returns the As the Executive switch, its six Ask first switches (`{kind, label, hint, on}`, the hint a plain line on what each one catches), the company's rules, whether this server can tie a change to the owner (`available`), and whether the Executive is paused. `PUT /take-the-lead` takes `{enabled?, ask_first?}`, and `POST /take-the-lead/rules` / `DELETE /take-the-lead/rules/{id}` add or remove a company rule (`{kind: person|domain|words|amount, value}`, 422 on a bad one). All four are the principal's alone (403). Anything that lets more happen on its own (turning it on, turning an Ask first off, removing a rule) needs a caller the API can tie to them (`delegation.verified.caller_refusal`, else 409). `GET /today/activity` adds `took_the_lead` rows (what Take the lead did, named, not quoted) and, for the caller alone, `sent_as_you` rows for what Handle it for me sent as them (and the meeting invites it accepted as them, \"Accepted Dana's Pilot review, Wed 21 Oct, 14:30\") and, for the principal, one `contacts_kept` row a day the contact keeper changed something (`detail` like \"2 added, 1 updated\"). Under `/delegation`: `PUT /delegation/take-the-lead {enabled}` is Take the lead as you (the owner's only for now; turning it on turns Handle it for me on), and `GET|POST /delegation/take-the-lead/rules` and `DELETE …/rules/{id}` manage the caller's own rules. `HandleItOut` gains `lead` and `lead_available`. A `take_the_lead_action` decision is approved or rejected through the existing `/decisions/{id}/approve|reject`. Training: the response also carries `training` (In training: everything waits unless allowed), `learned` (`{id, label, feature, example, uses, created_at}`: what it was allowed, the edit kept as an example by field) and `suggested` (`{key, label, approvals}`); `PUT` takes `training?` too, and leaving training while on needs the same provable owner as turning it on. `POST /take-the-lead/learned {key}` allows a current suggestion (404 for anything else; provable owner), `DELETE /take-the-lead/learned/{id}` makes it ask first again. `learned` also lists the owner's own Act as me senders (`feature` `act_as_me`), never anyone else's, and Remove works on those too. On a `take_the_lead_action` card, `POST /decisions/{id}/approve` takes `edits.input` (the card's editable fields, 422 for any other; the principal's alone, 403 for an area approver) and `edits.allow: true` (a training card only, 422 otherwise; the principal from a provable request, 403 otherwise), finishing `approved_with_edit` when a field changed.",+ "markdown": "The FastAPI surface lives in `openexecutive.api.main`, which mounts one router per topic. A shared-secret middleware gates every request via an `x-api-key` header, except `/health` and the `/webhook/*` callbacks (which carry their own verification). A request it lets through is marked `request.state.shared_secret_ok`, which is what lets an MCP request turn on \"may act\" with `x-mcp-client-actions: on` (`mcp_server.may_act`; with no secret the header is ignored and `MCP_CLIENT_ACTIONS` decides). With local login (`OE_LOCAL_LOGIN=1`, which only `make dev` sets, and only while neither Google nor SSO sign-in is set up) a second middleware answers 403 to any request whose raw `Host` is not `localhost`, `127.0.0.1` or `[::1]` (except a webhook that verifies its caller: Google Chat and Slack always, Telegram only with a `TELEGRAM_WEBHOOK_SECRET` Telegram can send), and to any write whose `Sec-Fetch-Site` is present and not `same-origin`/`none`. The UI signs that owner in with no password and this API usually has no secret, so without it a web page in the owner's browser could drive the API as the principal — by DNS-rebinding its own hostname to 127.0.0.1, or by posting a plain form from another site or localhost port. CORS is restricted to the UI origins in `BACKEND_ALLOWED_ORIGINS`, plus the `make dev` UI on `localhost:3000` only when `OE_PUBLIC_DEPLOYMENT` is unset, so a public instance never answers a localhost page with credentials. The gate is off when `BACKEND_SHARED_SECRET` is unset — intended for local development only — so any internet-reachable deployment sets `OE_PUBLIC_DEPLOYMENT`, and `create_app()` then refuses to boot without the secret rather than serving an unauthenticated API.\n\n**Who is calling** is read in one place, `api/caller.py`, which answers with a `Caller` kind and email; a test fails on any other read of an `x-caller-*` header. By default (`open`) it is the `x-caller-email` the UI proxy stamps from the sign-in, trusted because whoever holds the shared secret is trusted as the proxy, and a request with none runs as the principal. **Signed callers** remove that trust. With `CALLER_ASSERTION_PUBLIC_KEYS` set, the proxy signs every request with an Ed25519 key only the UI holds, and a third middleware checks the `x-caller-assertion` header:\n\n- **What it covers.** It is good for one method, the exact path and query, at most 120 seconds, and one use. The API keeps used ids in memory.\n- **What it names.** A `user`, by email, or the `operator`: local login, which runs as the principal.\n- **No caller header at all** is a `service`, for scripts or MCP clients holding only the secret. It is never the principal, not even on an install with no owner yet, so it can't save a first setup.\n- **Anything else is a 401.** That includes an `x-caller-email` without an assertion (`caller_assertion_required`), and a bad, expired, reused or mismatched assertion (`caller_assertion_invalid`).\n\nPublic keys that can't be read stop the boot, rather than fall back to trusting the header. `scripts/make-caller-keys.py` makes the pair, and `scripts/mint-caller-assertion.py` signs one request for curl. The API holds only public keys. Setup's API-protection check warns while a shared secret is set without them.\n\nEndpoints grouped by router:\n\n| Router | Key endpoints | Purpose |\n|---|---|---|\n| `chat` | `POST /chat`, `POST /chat/stop`, `POST /chat/add`, `POST /chat/background`, `GET /chat/voice`, `POST /chat/transcribe` | SSE stream of a live Executive turn, with a `: keepalive` comment every `CHAT_STREAM_KEEPALIVE_S` while the turn is quiet. `POST /chat/background` starts a new chat that the server answers to the end with nobody connected (a phone approving a card, then closing the app): it answers `202` with the new `session_id` once the same caller, access and quota checks as `/chat` pass, the turn saves itself as a `/chat` turn does, and one caller may have three running at once (`429` beyond that). `POST /chat/add` hands the running turn a message the person typed while it works; the turn takes it at its next step and the stream names it in a `message_added` event, and a second `POST /chat` on the same conversation waits for the first to finish. `GET /chat/voice` says whether the chat mic shows (`available`, `max_seconds`); `POST /chat/transcribe` takes one recording (multipart `file`) and answers `{text}` for the message box, nothing kept: `404` when speech to text isn't set up, `400` for a file that isn't audio, `413` over 25 MB, `502` when the speech server fails. The request body optionally carries `page_context` (route, page title, guide section id, and a form descriptor with current values) — sent only by the Ask OE side panel so the Executive can explain the current page or propose form values. When the Executive calls its `propose_form_values` tool, the stream emits a `form_patch` event (`{type, form_id, fields, rationale, iteration}`) that the panel applies into the on-screen form as highlighted suggestions; the user reviews and saves through the form's normal submit path. Each round of tool calls also emits an `activity` event (`{type, label, tool, iteration}`) naming what is running — progress only, forwarded verbatim and never persisted with the message, unlike an `action_taken` chip. A `session_id` in the body must be a chat the caller may use (the same ownership rule as `sessions`). If it isn't, the turn runs in a fresh chat instead, so nothing is written into someone else's session and the reply looks the same as for an unknown id. An unknown namespaced id (`slack:…`, `telegram:…` and so on) is ignored and a fresh chat is started, because those ids belong to the channel adapters and claiming one first would make the caller its owner. The client mints a `client_turn_id` and sends it with the turn; `POST /chat/stop` takes that id and halts the turn mid-flight, so the button is live from the moment Send is pressed rather than only once the first SSE byte arrives. A stopped turn emits a `stopped` event and then the usual `done` over the same still-open stream — never an `error` — and whatever streamed so far is persisted and flagged. The endpoint returns 404 both for an unknown id and for another caller's turn, so it cannot be used to probe which turns are live. An optional `memory_text` (1–2000 chars) is what peer memory records as the caller's words instead of `message` — the briefing handoffs send it because their seed quotes the Executive's own card; `POST /chat/upload` derives it from the typed text and filenames. Since it never reaches the transcript, the `chat_turn` audit row keeps it beside `message`. After the reply — a stopped or timed-out one too, before the `stopped` / `error` / `done` frames — a turn that looked at anything sends one `sources` event, `{type, session_id, sources: [{kind, title, url}], unavailable: [area]}`: `kind` is `company`, `knowledge`, `notion`, `research`, `document` or `web`, `url` is only ever an http(s) link or, for an earlier document, `/artifacts/<id>`, at most 12 sources (6 from the web, pages the reply cites first), and `unavailable` names parts of the analysis (\"finance\") whose specialist failed or said nothing every time it was asked. It is persisted with the reply as `chat_messages.sources`. When the turn held steps for an outside read that still wait on the speaker's go (`outside_content.pending_held`), it then sends one `held_steps` event, `{type, session_id, steps: [plain words]}` (not after a stop or a timeout), which the UI shows as Approve / Not now under the reply; Approve sends \"Go ahead\". `GET /sessions/{id}/messages` puts the same `held_steps` on the last reply while they still wait (never for a shared chat's member). A request that carries `x-caller-email` marks its turn signed in (`Session.web_caller_signed_in`), which Act as me needs on a deployment without local login. |\n| `sessions` | `GET /sessions`, `GET /sessions/shared-with-me`, `GET /sessions/topics`, `PATCH\\|DELETE /sessions/topics/{topic_id}`, `POST /sessions/topics/{topic_id}/merge`, `POST /sessions/topic-areas/rename`, `POST /sessions/topic-areas/ungroup`, `GET\\|PATCH /sessions/{id}`, `GET /sessions/{id}/messages`, `DELETE /sessions/{id}`, `POST /sessions/{id}/messages/{message_id}/feedback`, `GET /sessions/{id}/followup`, `GET\\|PUT /sessions/{id}/shares`, `DELETE /sessions/{id}/shares/me`, `GET /sessions/{id}/people`, `GET /sessions/{id}/live`, `POST /sessions/{id}/typing`, `GET /sessions/{id}/export`, `POST /sessions/{id}/rewind` | List, read and delete saved chat sessions (assistant messages carry their row `id` and any `feedback`). The list is scoped to the caller, and each chat carries `preview`: its last message as one plain line, Markdown marks stripped and cut at a word near 120 characters, with `You:` before the person's own words. With `?q=` it is a search: the caller's own chats whose title or message text contains the words (case-insensitive, `LIKE` with wildcards escaped), each with `match_count` (messages that contain them; 0 for a title-only match) and `snippet` (the newest matching message, around the match, prefixed `You:` or `Executive:`). Scoped exactly like the list, so nobody else's conversation can match. The Chats page searches this way and opens a result with `?find=`, so the chat starts on that message and steps through the others. Each of the owner's chats also carries `topic_id`, the topic it is filed under (`memory.chat_topics`). `GET /sessions/topics` lists the caller's topics (`id`, `name`, `chat_count`, `kept`, `pinned_at`, `area`, `updated_at`) and, when some of their chats have no topic yet or some topics no area, starts filing them in the background from their titles and answers `organizing: true` until done. `PATCH /sessions/topics/{topic_id}` renames (a name another of their topics has merges the two) or pins one, `POST …/merge` moves its chats into another (`into_id`), and `DELETE` ungroups it, leaving its chats under no topic; another person's topic is a 404. `POST /sessions/topic-areas/rename` (`area`, `name`) renames one of the caller's areas and `…/ungroup` (`area`) drops it, its topics standing alone; an area they don't have is a 404. `PATCH /sessions/{id}` also takes `topic_id` (null for none) or `topic_name` (made if new) to file the chat, which then is never refiled on its own. A chat-app conversation started with /new is left off the list until its first turn (`memory.conversations`). Every per-session route checks ownership, because session ids are guessable (`slack:dm:<user id>`, `telegram:<chat id>`): only the session's owner or the principal gets through, and a legacy row with no owner is the principal's alone. A teammate's own chat (a web chat, a DM, a private Telegram chat; not a channel or group conversation, `chat._shared_space`) is theirs alone, the principal included, as their mail with the Executive is. Anyone else gets the same 404 as for an unknown id, so the routes can't be used to find out which chats exist. The one other way in is having started the chat in this process (matched by Person id or verified email, so it survives the caller being added to the roster), which is how a caller the roster can't resolve keeps using their own chat. That record is in memory, so after a restart such an ownerless chat is refused for reads. The principal can continue someone else's ownerless chat without becoming its owner. Deleting a session also drops its in-memory copy, so a later turn naming that id starts fresh. The feedback route records 👍/👎 (`{\"feedback\": \"up\"|\"down\"|null}`) on one assistant reply; only the session's own caller or the principal may rate it. The rater is stored (`feedback_by_person_id`), and a 👎 on a reply to the rater's own message triggers a working-style pass for them (paced; see attunement). The chat stream's `done` event carries `message_id` for the persisted reply. The followup route returns `{\"suggestion\": string|null}` — one suggested next user message, written by the utility-fast model from the last few turns, that the chat composer shows as its greyed placeholder (Tab / → or the inline chip fills it in without sending). It is null unless the session ends on a persisted assistant reply, and null (uncached) when the model fails. Same ownership check as feedback; successful suggestions are cached in-process per (session, reply id). Messages from `GET /sessions/{id}/messages` carry each assistant reply's saved `sources` (the same shape as the chat stream's `sources` event, without `type`), so a reloaded chat shows them again. The owner (not the principal on someone else's chat; the principal for an ownerless one) also has the chat controls. `PATCH` takes `{title?, pinned?}`: a name they give is kept (`sessions.title_locked`, so the automatic title never replaces it) and a pinned chat lists first, newest pin first, with `pinned_at`. `PUT /sessions/{id}/shares {person_ids}` adds exactly those team members to it (not contacts, the archived or the owner: 422), audited `session_shared`; a chat that read its owner's own mail (`mail_private`) is never shared (409) and stops being open to those it was shared with before, and a chat-app conversation is never shared (409). Those added get it from `GET /sessions/shared-with-me` (with `shared_with_you`, `shared_by`, `shared_at`) and **write in it with the owner**: `POST /chat` with its id runs their turn as themselves (`chat._shared_member` widens the turn's access, never its ownership), and they may read it, its messages (without the owner's ratings) and its export. They can't rate, rename, delete, share or redo it, and `DELETE /sessions/{id}/shares/me` takes them out of it (audited `session_shared`, with them under `removed`). In a chat several people are in, messages carry `sender_person_id` and `sender_name` (a message from before it was shared is the owner's), `GET /sessions/{id}/people` returns `{you, people: [{person_id, full_name, owner}]}` (owner first; empty for a chat of one), and redo is refused (409) when the last message is someone else's. `GET /sessions/{id}/live` is a server-sent event stream each open page holds (`api.chat_live`, one in-process broker): `ready` first (with the running turn so far, if any), then `turn_started` (who sent what), the reply's `chunk` / `thinking` / `activity` / `action_taken` / `phase`, `turn_ended` (the page reads the chat again), `typing` (from `POST /sessions/{id}/typing`, at most once per 2 s per person) and `people`. `removed`, `deleted` and `resync` (a page that fell 1024 events behind) end it; the access check runs again on every 20 s keepalive. At most 12 pages per person, 40 per chat and 2000 per process (429). Archiving someone takes them out of every chat shared with them, and their open streams close at the next keepalive. The page closes it while its tab is hidden. `GET /sessions/{id}/export` returns the chat as a Markdown file (`memory.chat_export`: each message under its writer's People name or `Executive`, the reply's sources listed under it). `POST /sessions/{id}/rewind {last_message_id}` takes back the last turn for Try again and Edit: under the conversation's turn lock it deletes the last reply and the messages it answered, but only while that reply is still the chat's last row, and returns those messages' words (`{message}`), which the page sends again as an ordinary `/chat` turn through the usual gates. 409 while a turn is answering, for a stale reply id, and for a chat-app conversation (its reply also lives in Slack, Telegram or Discord); audited `session_rewound`. |\n| `today` | `GET /today`, `GET /today/activity`, `GET /today/activity/daily`, `GET /today/brief-delivery`, `GET /today/value`, `GET /today/top-three`, `GET /today/weekly-review`, `POST /today/seen` | Morning-brief aggregation (in the install's first week, for the principal, `first_week_card` `{needs_goal, findings: [{title, summary}], brief_tomorrow}` or null), recent-action rail, Pulse heartbeat heatmap (per-day activity counts), and the owner's month tally for Pulse's Done for you (`GET /today/value?period=month|week&offset=N`: counted lines beside the period before plus the finished reports, owner only, 403 otherwise; see Today). `GET /today` carries the viewer's cached \"What's going on\" `narrative` with `narrative_generated_at` (when it was written) and `narrative_stale` (a fresher one is being written — the page re-polls; see Today), the same text split for Home's Right now card (`narrative_lead`, `narrative_points` `[{text, since}]`, `narrative_move`), `narrative_seen_at` (this reader's last look, which `POST /today/seen` records and returns as `{seen_at}`) and, for the principal, `morning_brief` `{at, text}` and `calendar_next` `[{title, start, all_day}]`. A proposal backed by a roster request carries `roster_request` (the new sender's server-derived details and held-message previews; the principal's only) and is answered through the `people` routes below. `brief-delivery` returns the latest morning brief or end-of-day digest that didn't reach the owner, as `{brief, at, problem, fix, readable}` in plain language (`readable` is false when it couldn't be written, so there is nothing to read), or null — null while briefs go out, once the cause is fixed, while the scheduler is off, and for anyone but the owner (`chat._caller_is_principal_or_unclaimed`); it carries no address or channel id. `top-three` and `weekly-review` feed the solo Briefing's own cards and answer only in solo mode and only to the principal (`workflows._caller_is_the_principal`: a principal match or no caller header, never an unclaimed roster); anyone else, and team mode, get null. `top-three` returns `{items: [{key, kind, text, why, slot}]}` — the morning brief's top three from the same code (`build_top_three(principal_due_soon())`), `slot` a free block like `10:00–11:00`, `\"\"` when none is left, null when no calendar was read — and is kept off `GET /today` because its calendar read may take up to 4 s. `weekly-review` returns the latest completed `weekly_review` run as `{run_id, completed_at, period, top_three, excerpt}` (next week's top three as plain text; `excerpt` when that list is empty), or null when none has completed. |\n| `people` | `GET\\|POST\\|PATCH /people`, `POST /people/{id}/archive`, `GET /people/requests`, `GET /people/requests/{id}`, `POST /people/requests/{id}/approve`, `POST /people/requests/{id}/decline`, `GET\\|POST /people/{id}/open-loops` (a `POST` tells the owner once and returns `notified`: `sent` with the channel, `held`, `not_reached` or `skipped`; see Attunement), `POST /open-loops/{loop_id}/close`, `GET /people/{id}/outreach`, `GET\\|PUT\\|DELETE /people/{id}/attunement` | Roster CRUD and scope lookups. Every `Person` carries `kind` (`team` / `contact`). `GET /people` lists team members only unless `?include_contacts=true` (the People page asks for both; the department-head and workflow pickers keep the default); `POST` / `PATCH` accept `kind`, and a principal contact is a 422 either way. They also take `email_aliases` (other addresses the person writes from, at most 10; on `PATCH` the full list replaces the current one, omitted leaves it): an alias matches their mail and may be emailed but never signs in; a malformed one is a 422, and an alias — or a new primary `email` — that is already someone else's address or alias is a 409. Every `POST` / `PATCH` then closes any pending roster request whose sender now matches someone and replays its held messages. Adding, editing and archiving (`POST`, `PATCH`, `/archive`) are the principal's only, or anyone's while no one is principal yet so a first setup can add its owner (`chat._caller_is_principal_or_unclaimed`, the check behind `PUT /workspace`; a request with no `x-caller-email`, such as the CLI or local login, counts as the principal). Anyone else, a person editing their own row included, gets 403 before the person lookup, so it can't probe which ids exist. Contacts are private to the principal (`caller_is_principal`: the signed-in email is the principal's, or no `x-caller-email`; stricter than the owner rule above, so while no one is principal nobody is): for anyone else `include_contacts` is ignored, a contact's id is a 404 on `GET /people/{id}` and open-loops (exactly as for a missing id; `PATCH` and archive are already a 403 for them), and creating a contact or turning someone into one is a 403, even on a first setup with no principal yet. `GET /people/me` returns `{person_id, is_principal}` so the UI offers contacts to the principal only. `GET /people/by-scope/{token}` never returns a contact. Archiving a person closes their open loops. The open-loop routes list what a person owes (soonest due first) and close one (`reason`: done / not_needed / cancelled) — both the principal's or the owner's only (403 otherwise, via `people.store.is_principal_or_self`). `POST /people/{id}/open-loops` `{task, due_date?}` assigns that person a task (201 with the loop; `due_date` is a local date, due 17:00 in the user's timezone, clamped to 60 days, default 2 days): any active team member may assign anyone on the team, 403 for anyone else; 409 for an archived person, a contact (as seen by the principal), a person at the open-loop cap, a duplicate or open loops turned off; 422 for an empty task. Nobody is messaged at assignment — the owner is chased once it is due. `/outreach` returns, per kind of proactive outreach over 30 days, `sent` / `replied` / `acted` / `ignored` / `pending` (voided DMs are excluded) — same access rule, checked before the person lookup so it can't probe which ids exist. `/attunement` reads, replaces (`rules`: up to 4 strings, or omitted to only set `locked`) or resets the person's working-style rules — same access rule, 404 for an archived person; `PUT` rejects a rule that is not purely about style with 422. Archiving a person also drops their working style and frees their aliases. The **roster-request** routes (declared before `/people/{person_id}`) are the principal's alone (`caller_is_principal`; anyone else gets 404, as for a missing id): `GET /people/requests?status=pending|approved|linked|declined|expired|superseded|all` lists up to 100 (`id`, `channel`, `channel_ref`, sanitised `display_name`, `profile_email`, `on_company_domain`, `suggested_kind`, `suggested_person_id` / `_name`, `status`, `resolved_person_id` / `_kind`, `message_count`, timestamps, `ack_sent`, and `previews` of the held messages while pending — shown to the principal, never to a model); `approve` takes either `link_person_id` (+ `replace_channel_id`) or `full_name` + `kind` (no default: 422 without both) + optional `role`; `decline` takes no body. An answered request is a 409, a chat id or address already on someone else is a 409; the held messages are replayed (or dropped) in the background. |\n| `contacts` | `GET /contacts/changes?person_id=`, `PUT /contacts/keeper`, `POST /contacts/changes/seen`, `POST /contacts/changes/{id}/undo`, `POST /contacts/changes/{id}/accept`, `POST /contacts/changes/{id}/dismiss`, `GET /contacts/{person_id}/note` | The contact keeper (see Org), principal only (404 for anyone else). `GET /contacts/{person_id}/note` returns what is remembered about one contact or teammate (`note`, `topics`, `last_emailed_at`, `last_heard_at`, `last_met_at`, `next_meeting_at`, `next_meeting_title`, `emailed_count`, `met_count`, `note_written_at`, `updated_at`), 404 when there is nothing yet or they are not active. `GET` returns `{mode, status, last_run_at, changes}`: the switch (`on` / `ask` / `off`), how the last pass went (`ok`, `act_as_me_off`, `not_connected`, …) and the last 30 days of changes (each `id`, `person_id`, `action` add/update, `field`, `email`, `full_name`, `old_value`, `new_value`, `reason`, `evidence`, `status` applied/suggested/undone/dismissed, `seen`, `created_at`); with `person_id`, that person's every change. `PUT` sets the switch; `seen` is Got it; `undo` / `accept` / `dismiss` return the change, 409 when there is nothing to undo or decide. |\n| `chat-directory` | `GET /chat-directory/people`, `GET /chat-directory/channels`, `GET /chat-directory/slack/lookup?email=` | The Slack, Discord and Telegram people and rooms the bots can reach, by name, for the People and department pickers (`integrations/chat_directory.py`). Each answers `{slack, discord, telegram}`, every one `{configured, ok, options: [{id, name, detail, email}], truncated, members_available}`: `configured` false (and no options) when that bot has no token, `ok` false when the provider could not be read, `members_available` false when Discord refuses to list server members (the Server Members intent is off). `lookup` returns `{configured, match}`, the Slack person with that work email or null. Cached in-process for 60 s; tokens are never returned. The principal's alone, like changing the People list (`chat._caller_is_principal_or_unclaimed`); channels also open while no one is principal yet, people and lookup wait for one; a 403 otherwise, and the pickers then keep only the by-hand id field. |\n| `departments` | `GET\\|POST\\|PATCH\\|DELETE /departments`, `.../goals`, `GET\\|POST /goals`, `PATCH\\|DELETE /goals/{id}` | Department state and goals. A goal has an optional department and an optional `owner_person_id` (a non-archived team member, else 422). `GET /goals` lists every goal with `owner_id`/`owner_name` (its own owner, else its department's head, else the principal) and `department_title`. `POST /goals` takes a goal plus `area` (a name, ≤80): an existing department by slug or title, else a new one of that name (no specialist, as with `POST /departments`), else no department when blank. `PATCH /goals/{id}` edits any goal and may move it (`department_slug`, null = none; `owner_person_id`, null = follow the department) — only the fields sent change; `DELETE /goals/{id}` removes any goal. The department-scoped goal routes still reach only that department's goals. `DELETE /departments/{slug}` keeps its goals with no department, on their owners. `PATCH /departments/{slug}` also takes `watched_entities` (≤50 names, `[]` clears) — the entities the research watch policy may add watches for and route to this department. A `cadences` spec that does not parse is rejected with a 422 (empty means no cadence, i.e. the check-in is off); a changed `check_in` cancels the pending check-in and queues the next one on the new schedule. `GET` and `PATCH /departments/{slug}` return the department state plus `next_check_in_at` (ISO UTC of the pending check-in, `null` when it is off, the workspace is solo or none is queued); the list route does not carry it. A new `head_person_id` must be a team member: an unknown id and one of the principal's contacts get the same name-free 422 (`person_id N is not on the team.`), so the route — open to any signed-in user — never reveals that a contact exists; an unchanged head is not re-checked. `POST .../goals` requires only `key_result`: a missing or blank `period_value` is filled with the current period for `period_type` in the user's zone (the same label `create_goal` uses), and `target` defaults to empty. `PATCH .../goals/{id}` rejects a blank `key_result` or `period_value` with a 422; `target` may be cleared. |\n| `agents` | `GET /agents`, `GET /agents/models`, `GET /agents/models/options`, `GET /agents/presets`, `POST /agents/presets/{id}`, `PATCH /agents/{id}`, `POST /agents/{id}/test` | Specialist metadata and Council overrides. `/agents/models` is the flat model allowlist the PATCH validator uses; `/agents/models/options` returns the same ids in the same order, each tagged with `provider`, `provider_label`, `route` (`direct` / `openrouter` / `local`) and a display `label`, for the Council's Provider → Model picker. `AgentDetail` carries `instructions` (additional instructions, appended after `prompt`, which stays the built-in prompt or its replacement); `PATCH` and `/test` accept `instructions` (max 4,000 chars, blank clears it). `GET /agents/presets` lists the Fast, Balanced and Thorough quality presets (each with `available` and the resolved `model` for this install) plus `active` (null when the council is Custom), `base` and `custom_agents`; `POST /agents/presets/{id}` applies one by writing per-agent model and deep-reasoning overrides and returns the same shape (404 unknown preset, 409 when the install allows none of its models). Each `AgentMeta` carries `visibility` (`core` or `internal`) so the Council's simple view can list core agents first. |\n| `personas` | `GET\\|PUT\\|POST\\|DELETE /personas` | Executive persona definitions. |\n| `alerts` | `POST /alerts/{id}/approve`, `GET /alerts/approvals/{id}`, `POST /alerts/{id}/ack`, `POST /alerts/bulk-ack`, `POST /alerts/{id}/snooze`, `POST /alerts/{id}/reopen`, `POST /alerts/review` | Groom the briefing queue. `approve` (body `{edited_body?}`, 202) carries a proposal card out on the server and records who approved it and when (`alerts.approvals`, see Today): it queues the scheduled action a department-gated card held, or runs one Executive turn as the approver in a new chat (`session_id`). 409 for a card already approved or closed, a decision-backed card (approved at `/decisions`), a roster request or a drafted artifact; 403 unless the caller is the card's routed person or the principal (an unrouted card is the principal's); 429 when the caller already has three background turns running. `GET /alerts/approvals/{id}` reads one back (`running`, then `done` / `failed` with `outcome`) for its approver, the card's person or the principal (404 for anyone else). `snooze` (`until`, an ISO time up to 30 days ahead; `null` clears it, the Undo) hides an `unread` card until then by setting `snoozed_until` (409 on a closed card or a roster request); the TTL then counts from the snooze, so a card snoozed past its TTL still comes back. `ack` (`read`/`ack`/`dismissed`, optional `mute_topic`) also teaches a watch-sourced alert's watch (`trust_score`); `bulk-ack` takes explicit `alert_ids` and/or `older_than_days` + `category` (exempt sources skipped); `reopen` is the Undo for a `resolved` / `expired` / `dismissed` alert (never `ack` or an artifact / decision row; back to `unread` with every review stamp cleared and its TTL anchor refreshed, so it is live again even if older than its TTL); `review` runs the Executive's relevance review on demand (ignores the interval; still honours `ALERT_REVIEW_ENABLED`, the per-pass caps, the single-flight lock and per-alert idempotency) and returns the move counts. `bulk-ack` caps ids at 500, needs `older_than_days >= 1`, and applies watch feedback per row; `mute_topic` must be one of the alert's tags or a scoped `external:` / `department:` pattern. Alerts still surface only through the briefing (`/today` reads the live view) — there is no list, prefs, mutes, or live-stream endpoint. |\n| `artifacts` | `GET /artifacts`, `GET /artifacts/{id}`, `GET /artifacts/{id}/download?as=`, `POST /artifacts/{id}/archive`, `POST /artifacts/{id}/restore`, `DELETE /artifacts/{id}` | Browsable deliverables in any format — Markdown, HTML, Word, Excel, or a link into a connected app — with `format`, `downloads`, `external_url`, `supersedes_id` and, for an extension's kind of document, `collection` / `collection_label` on every summary (`orchestrator.extensions`; archive, restore and delete tell the collection first and answer 502 if it refuses). `download` renders the file (Word/Excel built at request time from the stored text; any Markdown artifact, workflow output included, can export `?as=docx`) and is always an attachment with `nosniff` and a sandbox CSP. Every route answers for the signed-in caller (`artifact_records.Viewer`): a drafted artifact is its owner's alone (`alerts.owner_person_id`, the person whose conversation published it; none = the principal's), a workflow output follows its run (`persistence.run_visible_to`), and anything the caller may not see is a 404 like a missing id, the principal included. `POST /alerts/{id}/ack` and `/reopen` answer 404 for someone else's drafted artifact the same way. |\n| `decisions` | `GET /decisions`, `POST /decisions/{id}/{approve,reject,cancel}`, `GET`/`PUT /decisions/classes/meeting_scheduling`, `GET /audit/reliability` | Approval-gated decision instances (e.g. calendar bookings). What approve/reject/cancel do is per class (`DECISION_CLASSES`, one `DecisionClassSpec` each; meeting booking books on approve and deletes on cancel; `delegation_reply`, an inbox reply card, is its approver's alone (`approver_only`) — hidden from everyone else, the principal included when it is a team member's — sends its Gmail draft on approve and, after a reject, deletes the draft unless it was edited; `delegation_invite`, a meeting invite card, is its approver's alone too, and answers the invite in their own calendar on approve (`edits.answer`) and nothing on reject; `department_action`, a team member's chat send or edit their department held (see Org), is the principal's and its approver's alone (`principal_and_approver`), makes that exact call on approve and audits `department_action_done` / `_failed` / `_declined`), and an instance of a class with no spec gets 409 instead of being carried out as another. Only the principal, or the person the decision went to for approval (`approver_person_id`), may approve, reject or cancel it. Anyone else gets 403, and a decision nobody was asked about, or one private to the principal, is the principal's alone. The resolver is recorded as `resolver_person_id`, and a reject that loses the race to another answer is a 409. `GET /audit/reliability` is the principal's alone (403 otherwise), because it counts private decisions too. A proposed booking is surfaced on the briefing as a companion alert; approve/reject/cancel here also clear that alert. An approval whose caller resolves to the principal runs the booking inside `grant_contact_egress()`, so a meeting the principal proposed with a contact can be created. Such a booking's payload is marked `private`: it is proposed to the principal, left out of anyone else's `GET /decisions`, and a 404 for them on `GET /decisions/{id}` and every action route. Reviewed in the briefing — there is no separate proposals page. The class routes read and set how meetings are handled: both return `{decision_class: \"meeting_scheduling\", mode: \"propose\" \\| \"auto_execute\"}` (default `propose`); the `PUT` body is `{mode}`, principal-only (403 otherwise, open before a principal exists — but then only `propose` can be set: `auto_execute` is a 409 until a non-archived principal is on the roster), 422 on any other mode, and a change writes a `decision_class_mode_changed` audit row. In solo mode the class mode decides a booking, and `auto_execute` applies only to a turn the principal started on a verified surface — any other turn gets a proposal. |\n| `delegation` | `GET`/`PUT /delegation`, `GET`/`PUT`/`DELETE /delegation/voice`, `POST /delegation/voice/learn`, `POST /delegation/voice/describe`, `POST /delegation/voice/signature`, `PUT /delegation/inbox`, `POST /delegation/inbox/check`, `GET /delegation/replies`, `PUT /delegation/replies/{id}/text`, `PUT /delegation/replies/{id}/recipients`, `POST /delegation/replies/{id}/answer`, `GET /delegation/actions`, `GET /delegation/invites`, `PUT /delegation/handle-it`, `PUT /delegation/disclosure`, `PUT /delegation/training`, `DELETE /delegation/learned/{id}`, `GET /delegation/handled`, `PUT /delegation/team` | Act as me — the **caller's own** delegation only; nobody sets it for anyone else. It answers the principal, and a team member once the owner lets them; anyone else gets 403 `not_available_yet`. `GET /delegation` returns `{enabled, gmail: {status, message, email, connect_command, outlook_connect_command, provider}, inbox: {enabled, status, message, watch_since, last_poll_at, checking}, handle_it: {enabled, mode, available, sent_today, lead, lead_available}, training: {replies, follow_ups, actions, drafts, learned}, disclosure: {enabled, handle_it, approved_sends, line}, team}`, where `team` is the owner's switch for the principal while `DELEGATION_TEAM_MEMBERS` allows it — `{enabled, members: [{person_id, name, enabled, inbox, drafts_30d, sent_30d}]}`, counts only — and `null` otherwise. `PUT /delegation/handle-it` takes `{enabled?, mode?}` (`mode` is `careful`, `balanced` or `bold`; anything else is a 422 `bad_mode`) from a caller the API knows is that person — 403 `not_yours`, 409 `caller_signing_required` with only the shared secret — and turning it on needs Act as me and Draft replies to my inbox on (409 `act_as_me_off` / `inbox_off`); `GET /delegation/handled` returns `{replies: [{decision_id, sent_at, to_name, to_email, subject, body, open_questions, gmail_link, source, summary}]}`, the caller's replies and follow-ups (`source` `follow_up`) sent on its own in the last 7 days, and the meeting invites it accepted on its own (`source` `invite`, `subject` the meeting's title, `summary` one line such as \"Accepted Dana's Pilot review, Wed 21 Oct, 14:30\"); a card on `GET /delegation/replies` carries `source` the same way, `can_allow` while Replies (Follow-ups, on a follow-up) is in training, and `learns_style` while Drafts is. `PUT /delegation/disclosure` takes `{handle_it?, approved_sends?}` (any other field is a 422), the caller's own switches for the AI line on replies sent as them (`disclosure.enabled` is the install's `AI_DISCLOSURE`); turning one on needs no signed sign-in, turning one off does (409 `caller_signing_required`), and a change is audited as `delegation_disclosure_changed`. `PUT /delegation/training` takes `{replies?, follow_ups?, actions?, drafts?, actions_on?}` (any other field is a 422; `actions_on` is Suggested actions On, and `actions` and `actions_on` set true clear each other) and returns the `GET /delegation` shape, whose `training` carries `actions_on` too; putting Replies or Follow-ups in, or Suggested actions out of training or On, needs no signed sign-in, anything else does (409 `caller_signing_required`), and putting Replies or Follow-ups in turns Take the lead as you off (`PUT /delegation/take-the-lead` on answers 409 `in_training` while they are). `training.learned` is `[{id, label, setting, example, uses, created_at}]`, what the caller allowed or kept in training (`setting` `replies`, `follow_ups`, `actions` or `drafts`; the kept example, or \"\"); `DELETE /delegation/learned/{id}` removes one of the caller's own (404 for anyone else's). `PUT /delegation/team` takes `{enabled}` from the principal alone (403 `principal_only`; 409 `not_available` when the install doesn't allow it). The Gmail `status` is `connected`, `not_configured`, `needs_reconnect`, `mismatch`, `no_email`, `shared_mailbox` or `error`, checked live with Google on each call. `PUT` takes `{enabled}` (any other field is a 422); turning it on needs the caller's own Gmail connected, otherwise a 409 whose code says why (`gmail_not_connected`, `gmail_needs_reconnect`, `gmail_mismatch`, `no_email`, `shared_mailbox`, `gmail_error`). The voice routes show, learn (from the caller's sent mail, one pass at a time — a 409 `in_progress`, `locked`, `too_soon`, `not_enough_mail`, `no_profile` or `changed` when it was locked or edited while learning, or a Gmail code), write it from the caller's own description (`POST /delegation/voice/describe` with `{description}`: one model call applies their words to the profile they have and returns it with a short `sample_reply`, saving nothing, so they review it and save it with `PUT`; a 409 `empty`, `too_long`, `in_progress`, `too_many` past 10 an hour or `no_profile`; no mailbox needed), edit and lock, and reset \"How I write\": a typed field that fails validation is a 422 `invalid_voice` listing each `rejected` field, and `clear_exemplars` / `clear_signature` drop those. `POST /delegation/voice/signature` reads the signature from the caller's Gmail settings again and changes nothing else (a locked profile stays locked), with the same Gmail codes as learning: a 409 when their mailbox isn't connected, it refuses the sign-in or the check itself fails (`gmail_error`), and a 502 `gmail_error` when reading the signature fails. On an Outlook mailbox it is a 409 `signature_unavailable`, since Outlook gives apps no signature to read, and the stored one is left alone. A caller `can_delegate` refuses — anyone but the principal, for now — gets 403 `not_available_yet`, which hides the Settings card. A request with no `x-caller-email` is a 403 `sign_in_required` unless the API runs under local login — a header-less call is not a sign-in (the caller is read through `api/caller.py`, as on every principal-only route). `PUT /delegation/inbox` takes `{enabled}` and turns **Draft replies to my inbox** on or off, returning the same shape as `GET /delegation`: on needs Act as me on (409 `act_as_me_off`) and the mailbox connected (the Gmail codes above, which keep their names for an Outlook mailbox too), and turning Act as me off turns it off too. `POST /delegation/inbox/check` starts a scan now in the background and answers 202 with the `inbox` state (409 `inbox_off` or `in_progress`). `GET /delegation/replies` returns `{cards: [...]}`, the caller's open reply cards from the database alone (no Gmail call): the decision id, the sender's name, address, relation and whether it was verified, the subject, when it arrived, what they wrote, a one-line `gist` (what they asked and what the draft says, written with the draft and brought up to date when an answer rewrites it; empty on an older card), the draft's recipients, subject and body, `open_questions`, `flags` and a Gmail link built from a fixed prefix. **Send** is `POST /decisions/{id}/approve`: it sends the card's draft exactly as it is in Gmail, after checking again that the caller is its owner — which on a server needs signed callers (409 `caller_signing_required` otherwise) — that both switches are on (`inbox_off`), no client slot is active (`client_slot`), the Gmail is connected (the Gmail codes), the draft is still there and unanswered (`draft_gone`, `you_replied`) and its recipients are allowed (`no_recipients`, `too_many_recipients`, `executive_recipient`, `not_from_you`). Recipients that changed, or a newer message in the thread, answer 409 `{code: \"confirm\", message, reasons, recipients}`; approving again with `edits: {recipients: <the list shown>, thread_moved_on: true}` is the second yes. Send with the *on its own next time* tick adds `edits.allow: true`: 422 `cant_allow` unless Replies (Follow-ups, on a follow-up card) is in training, 409 `learned_full` when their list is full, both before anything is sent; `edits.example: true` keeps a draft edited in the mailbox as how they write to that person while Drafts is in training. A draft edited in Gmail between those checks and the send is not sent (`draft_changed`). A double tap sends once (`already_handled`); Gmail refusing answers 429 `rate_limited` or 502 `gmail_error` with nothing sent, and a send Gmail didn't confirm answers 502 `send_unconfirmed` and is settled by the next inbox scan. Dismiss is `POST /decisions/{id}/reject`. `GET /delegation/invites` returns `{cards: [...]}`, the caller's own meeting invite cards (`delegation_invite`, Read my calendar on): `decision_id`, `status`, `created_at`, `from_name`, `from_email`, `relation`, `title`, `when` (in the workspace time zone), `start`, `end`, `all_day`, `clashes` (busy times only, `\"14:00–14:30\"`), `clash_checked`, `can_answer` (false for a sign-in that reads but can't answer the calendar; then `connect_command`), `can_allow` (Replies in training) and `waited_because`. **Accept** / **Decline** is `POST /decisions/{id}/approve` with `edits: {answer: \"accept\" | \"decline\"}` (accept is the default; `allow: true` with accept for the *Accept from them on its own next time* tick), on the same caller checks Send uses (403 `not_yours`, 409 `caller_signing_required`); 409 `calendar_write_needed` when the sign-in can't answer (the card stays), `gone` when the event left their calendar, `expired` once the meeting started, `already_handled`; 502 `calendar_error` leaves the card for another try. **Dismiss** is `POST /decisions/{id}/reject` and answers nothing. `GET /delegation/actions` returns `{cards: [...]}`, the caller's own approval cards (`delegation_actions`): the decision id, when it was made, a one-line `why`, and each action's `index`, `kind` (`invite`, `message`, `add_contact`), `summary` and `text` (a message as it will be sent, or a meeting's description). **Approve** is `POST /decisions/{id}/approve` with `edits: {only: [indexes]}` (and `allow: true` for the *Do this on its own next time* tick while Suggested actions is in training: 422 `cant_allow` otherwise or when nothing on it can be allowed, 409 `learned_full`; a card carries `can_allow`); it needs the card's own person signed in (403 `not_yours`, 409 `caller_signing_required`), Act as me on (409 `act_as_me_off`) and a card under a week old (409 `expired`), runs once (409 `already_handled`), and returns the decision with `final_payload_json.results`: per action `done`, `waiting` (sent to whoever approves meetings), `failed` with why, or `skipped`. Every change writes a private audit row. |\n| `history` | `GET /memories/history` (`?q=`), `PUT /memories/history/settings`, `PATCH`/`DELETE /memories/history/{id}`, `POST /memories/history/{id}/forget-conversation` | Always in the loop: the **caller's own** notes and switches only, resolved as for Act as me (a signed-in person on the roster, or the principal under local login; otherwise 403 `sign_in_required` / `not_on_roster`). `GET` returns `{notes, reply_notes, retention_days, effective_retention_days, company_retention_days, retention_choices, can_keep_notes, can_note_replies, can_set_company_retention, share_work_style, can_share_work_style}` (`can_keep_notes`: a team member on the People list may turn notes on; `can_note_replies`: they can use Act as me, so email replies are noted too; `share_work_style`: their own \"Share my work style with the team\" switch, `can_share_work_style`: a team member in a team workspace). `PUT` turns notes on only for someone who may keep them, and `share_work_style` on only for someone who may share it (403 `not_available` for either; turning either off always works), takes a person's retention only when shorter than the company's (422 `invalid_setting`), and the company retention only from the principal (403 `principal_only`). `PATCH` pins or corrects a note; `DELETE` forgets it; `forget-conversation` is \"Don't remember this\". A note that isn't the caller's is 404. Every change is audited privately to its person. |\n| `reflection` | `GET /memories/reflection` (`?person_id=`), `DELETE /memories/reflection/misses/{id}` | Learning: the turns the conversation review graded a miss (`memory.chat_review`), with the person's words, the reply, the tools run, what they wanted instead, the `kind` (`company` / `product`), the finer `cause` (null when the grade gave none) and the signals, plus `weeks` (`{week, graded, misses}` per ISO week, newest first) and `sees_everyone`. The caller is resolved as for `history` (403 `sign_in_required` / `not_on_roster`); each person sees their own, the principal those from chats `chat._session_access` lets them open (a teammate's own web chat or DM and a mail-private conversation stay their owner's; channels and groups are open), or one person's with `?person_id=` (a teammate naming someone else gets 403 `not_yours`); the weekly counts carry no words and are not filtered. The same response carries `lessons` (`memory/lessons.py`: live first, then retired, each with `shown`/`fine_after`/`miss_after` and `retired_reason`), filtered by the same rule, and `DELETE /memories/reflection/lessons/{id}` retires one (204; 404 for a lesson the caller may not see, already retired, or missing). `DELETE` forgets one miss (`is_principal_or_self`; 404 `not_found` otherwise) and writes a `chat_review` audit row. Each miss also carries its nightly replay (`memory/replay.py`): `replay_state` (`open` / `closed` / `evidence`), `replayed_at`, `closed_at`, `replay_notes`; and the response carries `replay` (`{misses, would_pass, evidence, last_run_at}`: of the caller's visible misses from the last 30 days, how many would go right today). |\n| `notifications` | `GET /notifications/settings`, `PUT /notifications/settings` | Settings → Notifications: the **caller's own** settings only, resolved as for the notes (a signed-in person on the People list, or the principal under local login; otherwise 403 `sign_in_required` / `not_on_roster`). `GET` returns `{quiet_hours: {enabled, start, end, days}, delivery, urgent_through, timezone, effective_timezone, held_count, gets_briefs, updated_at}` — `days` is `every_day` or `weekdays`, `delivery` `instant` or `digest`, `timezone` the zone they chose (`\"\"` for the workspace's), `held_count` how many messages wait for their quiet hours or brief (counted, never listed), `gets_briefs` whether they are the principal. `PUT` takes `{quiet_hours, delivery, urgent_through, timezone}` and replaces them (extra fields, equal start and end, a bad time or an unknown zone → 422); switching `digest` → `instant` makes what waited for the brief go now. Audited as `notification_settings_changed`. See Org. |\n| `audit` | `GET /audit/logs`, `GET /audit/logs/{id}`, `GET /audit/sessions/{id}`, `GET /audit/usage`, `POST /audit/log` | Durable audit trail of side effects. `/audit/sessions/{id}` adds a per-session token cost summary; `/audit/usage` aggregates token usage and OpenRouter cost across all sessions (totals, by day, by model, and by source — the part of the system that made the calls — including server-side web-search counts), plus `scripts`: how many `run_script` scripts ran (and worked, and in workflows), the tool calls they made, their time, and `turns_avoided` — an upper bound on the model turns those calls would otherwise have needed (each call past a script's first) — and `python_jobs`: how many Python jobs ran (and worked, ran a kept Custom tool, took a chat attachment), their time, CPU, largest peak memory and bytes in and out, and the tokens and charge of the turns that ran one. The Token usage page shows it as a Scripts section. Rows private to the principal (`private_to_principal`: a turn about their private mail, or a row naming one of their contacts on their own turn) are theirs alone: for any other caller, resolved as for `/alerts`, the usage totals leave them out. Once a principal is on the roster, anyone else reads only their own rows (a teammate's own turns, their Act as me work) in the list, by id and in a session, and the principal reads those as actions only — what was done, when and in whose conversation, with the text, details and `full` left out, and no text search over them; any other row's id is a 404, and a session shows only the rows the caller reads. |\n| `architecture` | `GET /architecture/sections`, `.../sections/{id}` | This documentation surface. |\n| `guide` | `GET /guide/sections`, `.../sections/{id}` | Plain-language user-guide content surface — mirrors `architecture` (static, version-controlled files; no LLM on the serving path). |\n| `health` | `GET /health`, `GET /health/honcho` | Liveness probes. Only `/health` is unauthenticated; `/health/honcho` sits behind the gate like everything else, because the exemption is an exact-path set. `/health` also carries `google_tools_missing`, the pinned Google tools startup priming could not find: `[]` once it has checked and they are all there, a list of names when a workspace-mcp rename or the configured services and tool tier left one out, and `null` for no answer — Google Workspace is not configured, or the probe has not finished or it failed. Only `[]` asserts they are present. It also carries `gateway`, the MCP gateway's state from its supervisor (`orchestrator.gateway_boot`): `state` is `off` (MCP not enabled), `no_servers` (enabled, the config names none), `starting` (an attempt is running), `up` or `down` (every attempt so far failed; `next_retry_at` is the next one), with `attempts`, `up_since` and `email_poller` (whether the poller that rides on the gateway runs); the failure's own text stays out, since the route is open to anyone, and goes to the log and the owner's notice on Home. A hosted instance whose gateway never came up used to answer `ok` here with no email going out or in. Only `[]` asserts the tools are present, so a probe that died cannot read as healthy; Setup status says which of the three `null` means. |\n| `version` | `GET /version` | The running version, the latest release on GitHub and `update_available`, for Settings → About. The GitHub answer is cached 6h, only a semver `tag_name` is used and the release link is built from it; `UPDATE_CHECK_ENABLED=false` skips the call and still returns the running version. No in-app updater: upgrading is host-side (`docs/deployment.md#upgrading`). |\n| `setup_status` | `GET /setup/status` | Settings → Setup status: one light per part of the install — AI model, company setup, owner, the Executive's address, API protection, Gmail, your own mailbox (Act as me), Slack, Discord, Telegram, Google Chat, the scheduler, the daily brief (when and where it goes) and Honcho — as `{checked_at, checks: [{id, label, state, summary, fix, link, last_activity}]}`, where `state` is `ok`, `warn`, `error` or `off` (optional and not set up), `fix` says what to do in plain language, `link` is an in-app page that helps (`/people`, `/onboard`) and `last_activity` is the channel's latest `integration_inbound` audit time. It tests live, with at most one free read-only call per service, each bounded to 5 s: Anthropic `models.list` (so an organisation key missing `ANTHROPIC_WORKSPACE_ID` shows up), Slack `auth.test`, Telegram `getMe` and `getWebhookInfo`, Discord `users/@me` only while the bot is still connecting, and — once the owner's own Gmail credential exists — Google's token refresh and Gmail `users/me/profile`. It also reads the listeners the lifespan leaves on `app.state` (Slack's socket, the Discord bot and its task, the MCP gateway), the scheduler heartbeat and the pause switch. When `google_workspace` is configured, the lifespan also starts `MCPGateway.prime_pinned_tools()` in the background (held on `app.state.mcp_prime_task`, cancelled at shutdown), so the pinned Google tools named in the prompt are callable without a search. With `DRIVE_SYNC_ENABLED` the lifespan also bootstraps the `drive_sync_scan` heartbeat beside Notion's (see Scheduler); the Drive folder sync has no HTTP route, only the `openexecutive sync-drive` / `purge-drive` CLI. With `CONFLUENCE_SYNC_ENABLED` it bootstraps the `confluence_sync_scan` heartbeat the same way; that sync is run by hand with `openexecutive sync-confluence` / `purge-confluence`. Sample values copied from `.env.example` are named as such without being sent anywhere. No token reaches a response or a log line: probes report a status code, a service's short error code or an exception's class name. Requests within 5 s of a run get that run's answer and concurrent requests share one run (kept per app on `app.state`), so the button can't become a stream of calls to those services. Gated like every other route; the UI adds a sign-in light of its own, from its own settings. |\n| `documents` | `POST\\|GET\\|DELETE /documents`, `GET /documents/{filename}`, `GET /documents/sources`, `POST /documents/sources/{drive\\|onedrive\\|notion}/sync`, `GET /documents/{drive\\|onedrive\\|notion}`, `GET /documents/{drive\\|onedrive\\|notion}/{id}` | Upload, list, view extracted text, and delete indexed company docs. `GET /documents` rows carry `source: \"upload\"` and the `domain` the upload was indexed under. The connected-source routes back the knowledge page's single Company documents list: `/documents/sources` gives each source's `{id, label, enabled, syncing, last_run, last_error, interval_minutes, file_count}`; `/documents/drive`, `/documents/onedrive` and `/documents/notion` list synced files from the sync state (`{enabled, last_run, last_error, files: [{id, name, url, modified_at, synced_at, indexed}]}`; a Drive row also carries `unreadable_reason`, why a file with `indexed: false` has no text — e.g. a scan OCR could not read — or null); `/documents/{drive\\|onedrive\\|notion}/{id}` returns the stored text (a OneDrive id is `<drive id>:<item id>`) for an id on record (the path comes from the state, never the caller). They are read-only — a synced file is removed where it lives. `POST …/sync` starts one tick of the same sync the scheduler runs as a background task (202); 409 when the source is not connected or a tick is already running (every sync holds a per-process run lock, so a manual run never overlaps the heartbeat; the route also keeps its own started task per source and re-checks just before starting one, so two simultaneous requests start one tick, not two), 429 with `Retry-After` within 60 s of the last successful run or of the end of the last tick in this process, failed ones included, so a failing or slow source can't be re-run back-to-back. The page's text view never loads images from synced content (a remote image would be a read beacon for whoever edits the shared folder). Upload is keyed on the filename, so re-uploading a document replaces its chunks rather than duplicating them, and `DELETE` removes the rows as well as the file. `domain` must be one of the nine specialist domains (including `sales`) or `general`; anything else is a 400 rather than a 200 that indexes where no specialist can retrieve. A scanned PDF is converted (`knowledge.pdf_reader`) before it is indexed and in the text view, as are scanned PDFs in `POST /chat/upload` and the client / onboarding intake uploads. The upload is streamed to disk 1 MB at a time and refused with a 413 once it passes 50 MB, and the file is parsed in a short-lived child process (see the RAG section), so a large document does not stay in the API process's memory. A Word or Excel file that cannot be parsed, or whose parse runs out of time, is a 422 with a readable reason. When every parser is busy, the response is a 503 with `Retry-After`. |\n| `executive` | `GET /executive/status`, `POST /executive/pause`, `POST /executive/resume`, `GET`/`PUT /executive/outside-content` | Operator pause switch for all autonomous work (scheduler, Gmail poller, workflow resumer); inbound chat and DMs stay live. Status returns `{paused, paused_at, paused_by, reason, held_actions, can_resume, name}` — `held_actions` counts due `pending` rows that fire on resume, and `name` is the Executive's name (`EXEC_DISPLAY_NAME`), which the web chat shows over its replies. `pause` takes an optional `reason` (≤200 chars) and is idempotent (re-pausing keeps the original start and reason); `paused_by` is the UI proxy's `x-caller-email`, else `api`. Behind the shared-secret middleware — not `SCHEDULED_ADMIN_TOKEN`, which the UI cannot send. Any caller may pause; `resume` is principal-only (`is_principal_or_self`, 403 otherwise; open while no principal is rostered; a header-less caller resolves to the principal as in `chat`), and status carries `can_resume` for the viewer so the UI hides Resume from everyone else. `GET /executive/outside-content` returns `{mode, can_change}` for the owner's **When I read outside content** setting (`ask`, `outsiders` or `trust`; `orchestrator.outside_mode`), and `PUT` with `{mode}` changes it for the principal only (403 otherwise), writing an `outside_content_mode_changed` audit row. |\n| `workspace` | `GET /workspace`, `PUT /workspace` | Solo/team mode, the user's time zone and the principal's role (`memory.workspace_settings`). Both return `{mode, timezone, effective_timezone, role_kind, role_title, reports_to, remit, measured_on, company_domains, company_domains_custom}` — `timezone` is what the user set (null when unset), `effective_timezone` the zone in effect (`timezone`, else `USER_TIMEZONE`, else UTC), and each role field is null until set. `PUT` is a partial update: only the fields sent change; `mode` is `solo` or `team` (not null), `timezone` an IANA zone, or null/blank to clear it; `role_kind` is `owner`, `in_house`, `independent` or `other`, and `role_title` / `reports_to` / `remit` / `measured_on` are free text (trimmed, at most 120 / 120 / 500 / 300 characters), each null or blank to clear it; an unknown zone, a bad mode or role kind, an over-long role field or an unknown field is a 422 and changes nothing. Switching to solo cancels every pending department check-in, switching to team re-schedules them, and a new zone re-times the pending briefs and reflection; a role change has no side effects (the next turn reads it). `company_domains` is the company's own email domains (a teammate matches by local part on them — see Org): `GET` returns the ones in effect (the stored list, or the domain of the principal's primary address, with `company_domains_custom` false) only to the same caller the role is shown to, `[]` for anyone else; `PUT` takes a list (at most 10 plain domain names, never a free-mail provider's — 422 otherwise, the value never quoted) or null / `[]` to go back to deriving them, and a change is audited with the new list. Each change writes a `workspace_settings_changed` audit row (a role change adds `details.role_fields_changed`, the names of the fields that changed — never their text). The role is the principal's: `GET` returns it only to a caller `PUT` would accept and reads it as nulls for anyone else. `PUT` is the principal's alone (`is_principal_or_self`, 403 otherwise; a header-less caller resolves to the principal as in `chat`) and open while no principal is rostered, so first-run setup can pick a mode. Both work before onboarding — unlike `company-profile`, nothing here needs a profile. |\n| `scheduled` | `GET\\|DELETE` scheduled actions (`?status=…&order=asc\\|desc`) | Pending and fired scheduled actions; `order` flips the `run_at` sort (asc=soonest-first for the upcoming queue, desc=most-recent-first for terminal-status history). `DELETE /scheduled/{id}` cancels a `pending` row (409 otherwise). Like pause it only stops work, so any signed-in user may — the UI proxy's `x-caller-email` is trusted when the request also carries the `BACKEND_SHARED_SECRET` `x-api-key` — as may a matching `X-Admin-Token` (`SCHEDULED_ADMIN_TOKEN`) or a loopback caller with no token set; otherwise 401 (token set) or 503. Audited as `scheduled_action_cancelled`. |\n| `watchlist` | `GET\\|POST\\|PATCH\\|DELETE /watchlist`, `POST /watchlist/describe`, `POST /watchlist/{slug}/approve`, `POST /watchlist/{slug}/decline` | External-monitoring watchlist + signals. `POST /watchlist/describe` `{text}` drafts a watch from a sentence (one forced tool call on the utility-fast model, `monitoring.describe`; the company profile's name, competitors, vendors and tickers ride in the user turn) and saves nothing: it returns the create fields (`slug`, `signal_type`, `target`, `display_label`, `trigger`, `cadence`, `mode`, `severity_floor`, `severity_ceiling`, `notes`) plus four plain lines (`watching`, `tell_you_when`, `how_often`, `how_loud`) written by code from those fields, after the create route's checks (registered kind, cadence, severity, the URL guard, the rss feed check, a free slug); 422 with a plain reason when it can't be a watch, 429 past 30 drafts an hour, 502/504 when the model fails. `GET /watchlist/{slug}` also returns `noise` `{recent_alerts, recent_dismissed, noisy}` (its last 8 alerts, noisy at 6 dismissed, `monitoring.noise`), and `PATCH` takes `quiet_noise_nudge: true` to stop that question for the alerts so far. `display_label` on create maps to the adapter's label key for every kind. Rows carry `origin` (`manual` \\| `executive` \\| `research` \\| `research_proposed`); `approve` turns a research suggestion (`research_proposed`, `dry_run`) into a live watch, `decline` `{reason: not_relevant \\| too_noisy \\| wrong_source}` removes it and records the decline (or, for `too_noisy`, keeps it live at a high floor), both 409 unless the row is a pending suggestion; `DELETE ?reason=` on a research-origin row records the decline too. |\n| `workflows` | `GET /workflows`, `POST /workflows/{name}/runs`, `GET\\|POST\\|PUT\\|DELETE /workflows/custom`, `GET\\|DELETE /workflows/custom/{name}/targets`, `POST /workflows/runs/{run_id}/decision`, `POST /workflows/runs/{run_id}/rerun`, `GET /workflows/tools/search?q=`, `GET /workflows/tools/describe?names=` | Workflow catalog, run management, and user-created (dynamic) workflow CRUD. `rerun` queues a failed scheduled run (a department check-in, a saved workflow on a cadence) once more, due now (see Scheduler); the principal may, and for a saved workflow the person it delivers to; 409 for a run that did not fail, one started from `/jobs` (\"Start this one again from its page\") or one already queued again. Custom definitions come back with a server-managed `owner_person_id` (the creator — the web caller on `POST`, set on insert only; a client-sent value is ignored), who is asked to approve a workflow's first write to a new target. `GET /workflows/custom/{name}/targets` lists the targets its tool steps may write to without asking and `DELETE …/targets?value=` forgets one. `POST /workflows/{name}/runs` is a 403 (\"Only the principal can run this workflow.\"), audited, before any run row exists, when the workflow is one only the principal may start in the workspace's mode (`weekly_review` always, `morning_brief` in solo) and the caller is not the principal. A request with no `x-caller-email` counts as the principal; with no principal on the roster, no signed-in caller does. This governs starting a run only. A run started by hand (this endpoint, or `run_workflow` / research in chat) records its starter as `owner_person_id` and is theirs alone: `GET /workflows/runs` lists the team's (scheduled, no owner) runs plus the caller's own and the ones waiting on them, and `GET` / `DELETE /workflows/runs/{run_id}` and `POST /workflows/runs/{run_id}/decision` are a 404 for anyone else's, the principal included. A signed-in email that is on nobody's People entry cannot start one (403), and a run whose starter can't be read is refused (503); its steps read and draft documents as the starter only. `POST /workflows/runs/{run_id}/decision` with `{decision: \"approve\"|\"reject\"}` answers a run awaiting a yes/no sign-off from the web app — 403 unless the caller is the person it waits on or the principal, 409 if it isn't awaiting or needs a written reply — through the same `apply_resolution` a chat reply uses. Custom create/update also rejects (422) any action-step tool that does not resolve, and so does `POST /workflows/custom/{name}/activate` when turning a workflow **on** — that click is the approval for a tool workflow chat saved switched off (turning off is never checked). Turning on a workflow with action steps requires the body `{is_active: true, definition}` carrying the definition the user reviewed; if the stored one differs (ignoring `is_active` and timestamps) it returns 409, and the switch itself is a compare-and-set on the validated revision (`dynamic_store.activate_if_unchanged`, keyed on `updated_at`), so a save that lands while the tool check awaits the gateway — even from another process — is a 409 too, so only what was shown gets switched on. `GET /workflows/custom` lists inactive definitions too; `GET /workflows/{name}` 404s for them. The two `tools` endpoints return `{tools: [{name, description, read_only, source}]}` for the advanced editor's tool picker and the review card; `read_only` is true only when the name says it reads (a label, not a control). The run stream adds a `progress` event (`step_id`, `summary`) while an action step uses tools. Each catalog entry (`WorkflowMeta`) carries `is_custom`, `background` (true for the system-run built-ins the UI files under “System”) and `quick` (true for a Quick workflow, one per playbook nothing else follows, named `quick_<playbook>`). The run endpoint streams SSE; its `awaiting_human` event carries `delivery` (`self` \\| `sent` \\| `suppressed` \\| `alerted` \\| `failed`) alongside `person_id` / `question` / `awaiting_until`, because the outbound guard or an unreachable approver means the question may never have been asked — a client must not render “waiting on them” for those. A paused run ends the stream with neither `done` nor `error`, so that `awaiting_human` frame carries `terminal: true` to say so explicitly, plus `resumable` — whether the run continues by itself once answered or stops at the gate. `GET /workflows/runs/{run_id}` returns the run row with the resume payload replaced by a small `resume_progress` summary (`gate_step_id`, `completed_step_ids`, or `null`): the payload holds every completed step's full text, and the run-detail page polls this endpoint while a run is unfinished. |\n| `knowledge` | `/knowledge/builtin`, `/external`, `/search` | MBA knowledge base and external sources. `/search` is the introspection parallel of `retriever.retrieve` and mirrors its filters — including the `general` catch-all on the company collection — so the panel cannot report a specialist seeing nothing that it does retrieve in chat. It accepts `general` in `domain_filter` for the same reason. `/search` applies the same review gate as `retrieve()` — withheld (pending or rejected) built-in, external and failure rows are dropped, so the panel never advertises knowledge the Executive cannot use; it degrades to unfiltered rather than failing if review state is unreadable. `POST /knowledge/builtin` refuses a name already held by shipped content, and `POST|DELETE /knowledge/failures` now register and deregister for review — without that a user-authored failure case study had no review row, so no SME decision could block it. |\n| `skills` | `/skills`, `/skills/search`, `/skills/{name}/restore` | Skill registry (the UI's **Playbooks**) CRUD and search. `PUT` on a built-in saves a company customization that shadows it; `DELETE` returns `{name, outcome}` — `deleted`, `reverted` (customization removed, built-in back) or `hidden` (built-in hidden for this company); `restore` un-hides; `GET /skills?include_hidden=true` lists hidden built-ins. List and detail carry `used_by` — `[{name, title, is_custom}]` of the workflows that follow the playbook (switched-off custom ones included) (`WorkflowMeta.playbooks` is the other direction). |\n| `evals` | `/evals/scenarios`, `/evals/runs` | Eval scenarios and run results. `POST /evals/runs` is a 403 when a workflow scenario names a workflow that only the principal may start in the mode it runs in (the weekly review; the morning brief in solo) and the caller is not the principal — the Jobs page's rule for starting a run; stored results are read under the existing `/evals/runs` rules. |\n| `memories` | `/memories/{decisions,initiatives,advice}`, `GET /memories/facts`, `POST /memories/facts/{fact_id}/{retire,approve,decline}`, `GET /memories/facts/approval`, `PUT /memories/facts/approval/{person_id}`, `GET /memories/people`, `GET /memories/people/{person_id}/conclusions` | Episodic-memory record management; the standing facts the Pulse Corrections tab reads (`{facts: [...], can_retire, retirable_ids, can_review}`; each fact carries `recorded_by_role` / `recorded_by_name`; `?include_inactive=` default true, honoured for the principal only; anyone else gets the active rows plus their own `proposed` ones, with the provenance (`session_id`, `turn_id`, `recorded_by_person_id`, `retired_reason`, and `source_quote` except on their own facts) blanked, never retired, replaced or declined text; `?limit=` 1–500) and their retire (the principal any fact, a teammate only one they recorded that the principal has not approved — else 403; the audit row names the teammate; 404 for an unknown or `profile` row, 409 when no longer active; writes a private `fact_retired` audit row); approve / decline a teammate's `proposed` fact (principal only, else 403; 409 when it is not proposed; private `fact_reviewed` row); and the per-teammate \"needs my approval\" switch (on by default; `GET` lists `[{person_id, full_name, needs_approval}]` for every active team member but the principal, `true` for anyone not yet set; `PUT` takes `{needs_approval}`, 404 for anyone else; both principal only; private `fact_approval_changed` row); plus the read-only peer-memory overview the Pulse People tab reads (`{status: ok|disabled|error, people: [...], conclusion_total}`; `?recent=` 1–50, default 5) and one person's full conclusion list, paged newest first (`{status, person_id, items, page, size, total, has_more}`; `?page=` ≥1, `?size=` 1–100, default 50; 404 when the person has no peer). Peer memory is each person's own: the overview holds only the caller's entry (none for a caller off the roster; `conclusion_total` counts only theirs), the principal included, and `/conclusions` for anyone but the caller is a 404. |\n| `review` | `/review/items`, `/review/stats`, `/review/trusted-defaults`, `POST /review/curate`, `POST /review/bulk-approve` | Knowledge review queue and annotations. Shipped knowledge registers as a trusted default (`approved` with a NULL `reviewed_at`), so the queue starts empty; `curate` (`{domain, action: start|stop}`, domain required) opts one domain in or out, and `trusted-defaults` returns domain → unreviewed count. `bulk-approve` takes a non-empty `domain`, `item_ids` (max 500) or `all_pending: true`, and 400s with no selector; it returns `{approved_count, item_ids}`. |\n| `onboarding` | `/onboard/interview/understand`, `/onboard/interview/start`, `/onboard/interview/message`, `/onboard/interview/draft`, `/onboard/interview/revise`, `/onboard/interview/{id}` (GET resumes, DELETE starts over), `/onboard/interview/commit` | Conversational company setup. The `onboarding_interviewer` agent asks up to four quick questions (a question response carries `question_options`, suggested answers to tap), then emits a draft profile + people + `draft_goals` (each `{title, target, owner_name, department_title}`) + departments only when the user listed people by department. **Nothing is persisted until `POST /onboard/interview/commit`** — the interview itself never writes. `start` is multipart (description + optional deck/one-pager uploads). `revise` changes one earlier answer (0 = the description) and runs the interview again from there. `commit` takes the reviewed `goals` too (422 when one names a person or department not in the draft) and saves them with owners, additively. It answers 403 when the draft would make a different person the owner and the caller is not the current owner (setup demotes the old principal, and any signed-in user can open it). It also takes an optional `owner_email` — the principal's sign-in email, which the review screen pre-fills from the signed-in user — checked before anything is written (422 if it is malformed, already another person's, or would replace the principal's existing email — setup only fills in a missing one; 403 if the caller is not the owner and it is not the address they signed in with, since the roster is the sign-in allow-list) and saved on the principal's People row, so the owner's chats resolve to them from the first turn. |\n| `workflow_designer` | `POST /workflows/designer/start`, `POST /workflows/designer/edit`, `POST /workflows/designer/message`, `POST /workflows/designer/draft`, `GET /workflows/designer/{session_id}` | The conversational New workflow wizard, and editing a saved workflow by conversation. `/edit` (`{name}`) opens a session on a saved custom workflow with no model call: the stored definition is the starting draft, and every response then carries `editing` (its name) and `original` (the stored definition as the session opened on it); drafts keep that name and its on/off state whatever the model writes (404 for an unknown workflow). The UI saves one with `POST /workflows/custom/{name}/save-edit` (`{session_id, definition}`): the server saves the session's own draft, never a client-sent body, and only when it equals the `definition` the user reviewed (a later message is a 409, so the click saves only what was shown); it re-validates it, keeps the stored on/off state, and writes only if the stored workflow is still the version the conversation opened on — otherwise 409, as it is for a session about another workflow or one with nothing drafted yet. It sits under `/workflows/custom` so anything that limits writes to custom workflows covers it too. The `workflow_designer` agent asks clarifying questions (`phase: \"question\"`, with `hint` and up to four suggested `options`), then returns `phase: \"draft\"` with `draft.definition` (exactly the body `POST /workflows/custom` takes), `summary`, and `assumptions`; every response carries the `transcript`. **Nothing is persisted here** — sessions are in memory, and the UI saves through `POST /workflows/custom`. Errors are fixed strings (422 over the 8,000-char message cap or the transcript cap, 404 for an expired session, 409 while another turn on the same session is still running, 502/504 for model failure or timeout). Mounted ahead of the `workflows` router. |\n| `python_jobs` | `GET /python-jobs/{job_id}/{name}` | A result file a Python job (`run_python_job`) wrote, the principal's alone (403 otherwise), served as a download (`Content-Disposition: attachment`, `Content-Security-Policy: sandbox`, `nosniff`, `Cache-Control: no-store`); 404 for a bad id or name or a file past its 7 days. |\n| `saved_tools` | `GET /saved-tools`, `GET\\|PUT /saved-tools/{name}`, `POST /saved-tools/{name}/rollback` | The scripts the Executive saved as tools (Settings → Advanced → Saved tools), the owner's alone (403 otherwise). The list gives each tool's name, description, on/off, current version, `kind` (`script` or `python`, a kept Python job) and the tools it calls, plus whether `SAVED_TOOLS_ENABLED` is on; the detail adds the script, every version and the last 20 runs. `PUT {enabled?, workflows?, version?}` turns one on or off (off also turns it off for workflows, so turning it back on never re-arms unattended runs), and turns the named `version` — the one the owner looked at, never whatever is current by then — on for workflows (or workflows off); a Python tool never turns on for workflows (409) — turning workflows on lets more run unattended, so it needs a request tied to the owner (signed sign-ins or local login, as Take the lead's switches do; 409 otherwise); `rollback {version}` makes that version the one that runs. Both are audited as `saved_tool_changed`. The Executive saves tools itself through `run_script`; there is no create route. |\n| `onboarding` (fallback) | `/onboard/start`, `/onboard/answer`, `/onboard/status/{id}` | The original 12-step form wizard, kept as a fallback (UI at `/onboard?mode=form`) and used by the `openexecutive onboard` CLI. Its final answer saves the profile and adds people with their emails and a principal, so once there is an owner only they can send it (403, and the session rolls back so the owner can still finish it). `/onboard/start` reads the workspace mode once for the session: solo skips the Team Size, Team Members and Fractional Executives steps. Each status response's `current_step` / `total_steps` count only the steps that session asks, and `optional` says whether the current step can be skipped. |\n| `company-profile` | `GET\\|PATCH /company-profile` | Structured company profile. |\n| `alerts` (private cards) | `POST /alerts/{id}/ack`, `/reopen`, `/alerts/bulk-ack` | An alert tagged `private:principal` (see Today) is a 404 on ack and reopen for anyone but the principal, and their bulk sweep skips it. A roster-request card (source `roster_request`) is a 409 on `ack` even for the principal — it is answered at `/people/requests/{id}`, since acking would leave the sender waiting with nothing to show it — and, as an exempt source, `bulk-ack` and `reopen` skip it. |\n| `auth` | `GET /auth/allowed-emails` | Non-archived **team members** with an email, offered to the UI's sign-in gate — a contact is never on it, so adding a client to reach them by email no longer lets them sign in. The gate takes the **union** of this roster and its own `ALLOWED_EMAILS` env list, so env entries always work (and are checked first, short-circuiting the fetch) and roster entries are additive — a roster change can never revoke a configured operator. With local login (`make dev` with Google sign-in not set up — the UI then listens on 127.0.0.1 only) the UI signs in without an email, sends no `x-caller-email`, and the backend resolves the caller to the principal, as it does for the CLI. |\n| `fixtures` | `/fixtures`, `/fixtures/{name}/load` | Demo-company fixture load/reset/snapshot. `POST /fixtures/reset` leaves no departments, like a new install (`departments_seeded` is always 0). |\n| `clients` | `/clients`, `/clients/generate`, `/clients/cockpit`, `/clients/save`, `/clients/{slug}/activate` | Client-company slots — multi-client mode for fractional use; create from current/blank/AI-generated intake draft (pasted notes and/or file attachments, multipart on `/clients/generate`), save-back switching, engagement metadata (PATCH), cross-client cockpit, delete. See the Client Companies section. |\n| webhooks | `POST /webhook/telegram`, `POST /webhook/google-chat`, `POST /webhook/slack` | Inbound integration callbacks (self-verified). `/webhook/slack` answers 503 unless `SLACK_SIGNING_SECRET` is set. |\n| MCP server | `POST\\|GET /mcp` | Open Executive exposed as an MCP server (Streamable-HTTP). Not a router — a mounted FastMCP sub-app. Same shared-secret gate (`x-api-key`); exposes company context as resources and the specialist council as tools, answered as the verified caller (`api.caller`) by the web app's rules. See the MCP Server section. | A card the Executive left for something asked in chat (`app_action`, `orchestrator/app_control.py`) is its person's alone too: its payload holds `action`, `id`, `value`, `title` and the `card` line shown; **Confirm** is `POST /decisions/{id}/approve`, which runs the action through the route its button calls with that request (a refusal comes back as that route's status and words, and the card is marked `failed`), and a card older than a day is turned down with 409.\n\n**What it does for you** (`api/routes/may_do.py`): `GET /what-it-may-do` is read-only. It returns `{paused, sections}`, three sections in order (`as_you`, `executive`, `departments`), each `{key, title, available, unavailable, notes, rows, footer, goes}`, and each row `{id, what, outcome, then, why, setting, items}`: `outcome` is `goes`, `waits`, `always_waits`, `drafts` or `off`, and `setting` names where to change it (`act_as_me`, `handle_it`, `take_the_lead`, `pause`, `department:<slug>`, or empty). `as_you` is the caller's own (`available` false, with why, when they can't have Act as me); `executive` is the principal's alone (`available` false for anyone else); `departments` is everyone's. Any team member on People may call it; a contact, someone not on People, or a request with no caller outside local login gets 403. See Delegation → What it does for you.\n\n**How far it goes as you** (`api/routes/delegation.py`): `PUT /delegation/reach` takes `{step}`, one of `off`, `when_asked`, `inbox`, `careful`, `balanced`, `bold`, `everything`, and answers the caller's delegation state as `GET /delegation` does. Going down turns the switches above the step off, highest first; going up runs each switch's own route in order (`PUT /delegation`, `PUT /delegation/inbox`, `PUT /delegation/handle-it`, `PUT /delegation/take-the-lead`), so the first refusal (409 `gmail_not_connected`, a signed sign-in missing, `in_training`, 403 for Everything when not the owner) stops it at a whole step. See Delegation → How far it goes as you.\n\n**Take the lead** (`api/routes/take_the_lead.py`): `GET /take-the-lead` returns the As the Executive switch, its six Ask first switches (`{kind, label, hint, on}`, the hint a plain line on what each one catches), the company's rules, whether this server can tie a change to the owner (`available`), and whether the Executive is paused. `PUT /take-the-lead` takes `{enabled?, ask_first?}`, and `POST /take-the-lead/rules` / `DELETE /take-the-lead/rules/{id}` add or remove a company rule (`{kind: person|domain|words|amount, value}`, 422 on a bad one). All four are the principal's alone (403). Anything that lets more happen on its own (turning it on, turning an Ask first off, removing a rule) needs a caller the API can tie to them (`delegation.verified.caller_refusal`, else 409). `GET /today/activity` adds `took_the_lead` rows (what Take the lead did, named, not quoted) and, for the caller alone, `sent_as_you` rows for what Handle it for me sent as them (and the meeting invites it accepted as them, \"Accepted Dana's Pilot review, Wed 21 Oct, 14:30\") and, for the principal, one `contacts_kept` row a day the contact keeper changed something (`detail` like \"2 added, 1 updated\"). Under `/delegation`: `PUT /delegation/take-the-lead {enabled}` is Take the lead as you (the owner's only for now; turning it on turns Handle it for me on), and `GET|POST /delegation/take-the-lead/rules` and `DELETE …/rules/{id}` manage the caller's own rules. `HandleItOut` gains `lead` and `lead_available`. A `take_the_lead_action` decision is approved or rejected through the existing `/decisions/{id}/approve|reject`. Training: the response also carries `training` (In training: everything waits unless allowed), `learned` (`{id, label, feature, example, uses, created_at}`: what it was allowed, the edit kept as an example by field) and `suggested` (`{key, label, approvals}`); `PUT` takes `training?` too, and leaving training while on needs the same provable owner as turning it on. `POST /take-the-lead/learned {key}` allows a current suggestion (404 for anything else; provable owner), `DELETE /take-the-lead/learned/{id}` makes it ask first again. `learned` also lists the owner's own Act as me senders (`feature` `act_as_me`), never anyone else's, and Remove works on those too. On a `take_the_lead_action` card, `POST /decisions/{id}/approve` takes `edits.input` (the card's editable fields, 422 for any other; the principal's alone, 403 for an area approver) and `edits.allow: true` (a training card only, 422 otherwise; the principal from a provable request, 403 otherwise), finishing `approved_with_edit` when a field changed.","mermaid": null,- "generated_at": "2026-10-21T00:00:00Z"+ "generated_at": "2026-10-21T16:00:00Z"}diff --git a/packages/core/openexecutive/architecture/prebuilt/org.json b/packages/core/openexecutive/architecture/prebuilt/org.jsonindex fbfaabc0..028f0952 100644--- a/packages/core/openexecutive/architecture/prebuilt/org.json+++ b/packages/core/openexecutive/architecture/prebuilt/org.json@@ -1,7 +1,7 @@{"section_id": "org","title": "Org Structure (Departments & People)",- "markdown": "The org model has two persistent registries backed by SQLite (both share `episodic_memory.db`): **departments** and **people**. A new install starts with **no departments**: nothing is created at boot. Setup's commit creates only the departments its draft names (none for a solo workspace or a team that names none), and the Executive's `create_goal` tool creates one when a goal names an area that has none. A department named after one of the eight catalogue areas in `departments/charters` (Strategy, Finance, People & Talent, Legal, Operations, Marketing, Product, Board & Investor Comms) starts from that area's charter, and setup and `create_goal` wire it to the matching specialist (at most one department per specialist). Workspaces created before this keep the departments they have. `seed_default_departments` still inserts the full catalogue once per database (tracked by a `default_departments_seeded` sentinel in `departments_meta`), but only tests and tooling call it.\n\n### Company setup (onboarding)\nBoth registries are first populated by **conversational onboarding** at\n`/onboard`. The user describes their business in their own words (optionally\nattaching a deck or one-pager); the `onboarding_interviewer` agent — Council-\nvisible but outside `SPECIALIST_REGISTRY`, so the Executive cannot reach it via\n`consult_specialist` — asks up to 5 clarifying questions, then emits a draft\ncompany profile, a leadership roster, and a department list. The user reviews\nand edits that draft inline (the same section editors `/company-profile` uses)\nbefore anything is written. Until a profile exists the owner (the principal)\nis sent to `/onboard` from Home and every shell page except Settings, unless\nthey chose \"I'll do this later\"; after the save and the voice step a last\nscreen shows what was kept and the two accounts worth connecting (the\nExecutive's own mailbox, and the owner's own for Act as me), and Home's\nbanner slot carries a getting-started checklist (profile, first question,\nthe two mailboxes, the team) until every required row is done or it is hidden\n(`ui/src/lib/firstRun.ts`).\n\nTwo properties matter here. First, **the interview never writes the profile**:\nthe exchange lives in a TTL'd session (2 h, 50 at most) that the API process\nholds and also writes through to the `onboarding_interviews` table of the\nepisodic SQLite database after every turn (`onboarding/interview_store.py`),\nso a restart or another worker finds it again (a newer stored row replaces\na held copy, so two workers never overwrite each other) and the browser,\nwhich keeps only the session id, resumes at the same question or draft via\n`GET /onboard/interview/{id}`. A session answers only to the caller who\nstarted it; anyone else gets the 404 an unknown id gets. Every answer stays on screen with an Edit link, on the draft review too: `POST /onboard/interview/revise` replaces one earlier answer (an edited description keeps its attached files), drops everything after it, resets the question count to what was asked before it, and runs the interviewer again, so it asks again from there or redrafts. A failed turn leaves the session as it was. Start over (on the \"Picking up where you left off\" banner) deletes the session at once with `DELETE /onboard/interview/{id}`. The single profile write happens at\n`POST /onboard/interview/commit`, so a rejected save leaves the draft editable\nand retryable, and a successful one deletes the stored transcript at once (it\ncarries the company's financials). Second, **the department step is additive** — drafted\ndepartments are matched onto existing ones by slug or title and updated in\nplace, unmatched ones are created, and nothing is ever deleted. That is\ndeliberately unlike the fixture loader's `_seed_departments`, which DELETEs\nevery row first: doing that here would drop departments the workspace already\nhas, with their goals and `specialist_key` wiring, on a re-run of setup. A\ndrafted department that matches nothing is created, wired to the specialist\nwhose area it is named after (`specialist_key_for_area`) unless another\ndepartment already has that specialist. Matching goes through\n`departments.store.match_department` (slug as given or as the store would slug\nit, else case-insensitive title), the same rule the `create_goal` chat tool uses. `profile.org_structure` is derived from the reviewed\npeople and department tables at commit time, so it cannot drift from them.\n\nExactly one person is marked `is_principal` and receives `WILDCARD` authority;\nno contact details are imported, even if the user volunteers them, because\nchannels and handles are a deliberate action on the People page. The one\nexception is the owner's own sign-in email: the review screen has a separate\nfield for it, pre-filled with the current owner's email on a re-run and with the\nsigned-in user's on a first setup. The commit checks it before writing anything\n(malformed, already another person's — as their address or one of their\naliases — or different from the owner's existing\nemail is a 422 — setup only fills in a missing one) and then saves it on the\nprincipal's row. One case is adopted rather than rejected: when the address is\nthe one the caller signed in with and sits on the current principal's row\nunder another name (a hosted sign-up seeds that row from the address when no\nname was given), `adopt_owner_row` renames that row to the drafted name\nbefore the roster upsert, so the first save updates the owner's entry instead\nof failing with \"use the owner's name exactly as on the People page\". Without it the signed-in owner matched no Person,\nso their chat history stayed empty until they found the People page. With\nlocal login (`make dev` without Google or SSO sign-in) the field is optional: that\nsession carries no email, which the backend already reads as the principal.\nBecause the save demotes the old principal, only the current owner may draft\nsomeone else as \"This is me\" when setup is re-run (403 otherwise); a re-run\nthat keeps the owner, or a first setup, is open to anyone who can sign in. On such a re-run, someone who isn't the owner may fill in the owner's missing email only with the address they signed in with (403 otherwise): the roster is the sign-in allow-list, so any other address would let them sign in as the owner. That keeps the way in for an owner whose entry has no email, which is also where Setup status and the Executive send them, since the People page refuses anyone it can't recognise as the owner. Re-running setup stays open to everyone deliberately: the Company Profile and Departments pages already let anyone signed in change the same things. The original\n12-step form wizard remains available at `/onboard?mode=form` and backs the\n`openexecutive onboard` CLI. It adds people with their emails, so once there is an owner only they can save it.\n\nA first setup opens on one free-text box: the user describes their work in their own words, optionally attaching a one-pager. `POST /onboard/interview/understand` reads that text once (no session, nothing stored) and reports what it already says: Personal or Team, the principal's role (owner / founder, an executive inside an organisation, independent or fractional, or other), title, who they report to, the company and this year's focus. The next screen shows those for a quick confirm, with Personal / Team as the one required answer; confirming writes `PUT /workspace` (mode, the browser's time zone and, for Personal, the role) and then starts the interview with the same text, so it asks only about what is still missing, at most five questions. If the read fails the text is kept and the user simply picks. A re-run over an existing profile skips these steps; they are changed in Settings. The tone question comes last, after the draft is saved. In solo the interviewer's first user turn carries a short\nhint, one static constant per role kind (`onboarding.interview.solo_hint`):\nalways draft only the principal, as principal, and no departments; an owner is\nasked about the offer, customers, pricing, cash and runway; an in-house\nexecutive about their organisation (the company profile is their employer),\ntheir function and remit, their manager, their team size as context rather\nthan a roster, and what they are measured on; an independent about their\npractice, clients and pricing; with no role given, the role-neutral hint asks\nwhich of these they are. Each also asks their top goals. The hint never\ncarries the principal's own text, and it rides in the user turn, so the cached\nsystem prompt is the same constant in both modes. The solo review screen shows only **You**\n(name and role) and the sign-in email, and commits that one principal with no\ndepartments — the existing departments stay, as the areas goals are grouped\nby. The form wizard (and the CLI) read the mode when a session starts and skip\nthe Team Size, Team Members and Fractional Executives steps in solo; each\nstatus response numbers the steps it actually asks and flags whether the\ncurrent one is `optional`.\n\n### Departments\n**Goals stand on their own.** A `Goal` has an owner and, optionally, a department (`department_goals.department_slug` is nullable; `owner_person_id` is nullable too). Its owner is resolved at read time (`departments.goals.effective_owner_id`): its own `owner_person_id`, else its department's head, else the principal, so a goal that follows its department moves with a new head. A goal no department check-in grades — one with no department, or in a department with no specialist — is graded weekly by the `goal_review` scheduled kind (`departments.goal_review`, Monday 09:00 in the user's zone, both modes) with `cso` (`goals.FALLBACK_REVIEWER`), through the check-in's own verdict contract and `persist_goal_verdicts`; it skips goals graded within six days and not edited since, so a quiet week makes no model call. Deleting a department keeps its goals: each with no owner of its own takes the department's head as owner, then all lose the department (`store.delete_department`). An old database is migrated in place (`_migrate_goals_stand_alone`: the table is rebuilt once to drop the NOT NULL, then the owner column is added). `GET /goals` lists every goal with `owner_id`, `owner_name` and `department_title`; `PATCH`/`DELETE /goals/{id}` reach any goal and can move it between owners and departments (an owner must be a non-archived team member, never a contact).\n\nEach department carries a charter, `Goal` rows, `ChecklistItem`s, and named cadences. Cadence specs (`daily@HH:MM`, `weekly@DOW@HH:MM`, `quarterly@DD-HH:MM`, UTC) run a `department_check_in` workflow when due (never in solo mode, below), then chain the next occurrence. A new department starts with `check_in: weekly@mon@09:00` (`DEFAULT_CHECK_IN_CADENCE`; it was daily, and departments created before keep theirs). An empty spec turns the check-in off. `PATCH /departments/{slug}` rejects a spec that does not parse with a 422, and a changed `check_in` re-times the department at once (`cadence.reschedule_check_in` cancels the pending check-in and queues the next one on the new schedule, or none when it is off). The department page offers Daily / Weekly (day and time) / Off rather than the raw spec, and shows the next check-in from `next_check_in_at`. The check-in is skipped (and still chained) when the department has no specialist, no Goals, or nothing new since its Goals were last reviewed (a row private to the principal does not count as new: whether it runs is visible to the department). An idle initiative is chased with the department's head, or with the principal when there is none. A department may map to a specialist agent via `specialist_key`, or stay informational. No department is seeded for the `sales` specialist (existing installs keep their departments unchanged), so its consults skip the department-memory prefetch and the Honcho department sync, as `triage`'s do. An assigned head gets a generated department-head voice persona.\n\n**Goals from chat.** Besides the goal routes and the UI, the Executive keeps goals current from a chat turn: `update_department_goal` records reported progress, and `create_goal` starts tracking a new goal the principal states (\"20 paying clients by the end of Q4\"). Neither needs a department: `list_department_goals` without `department_slug` lists every goal with its owner and department, `update_department_goal` names a goal by `goal_id` alone (a `department_slug` it is given must still match), and `create_goal` leaves `area` out unless the principal names one, taking an optional `owner_person_id` (a team member, from `list_people`). Only the principal, on a surface that verified it is them (`people_tools.is_principal_on_verified_surface`, the roster tools' rule: the web chat, their own Slack or Discord, a private Telegram chat with a valid webhook secret), may call it: a goal renders in every turn's org block as one of the principal's own, so a goal from an inbound email, Google Chat, a teammate or an unattended run would be text carrying their authority. Anyone else gets a refusal (audited with the caller and surface) and nothing is written. `create_goal` names the area as free text — a slug or a title, resolved by `match_department` — so its schema is static and adding an area never changes the cached tool prefix. When nothing matches it creates the area (`create_department`), giving it a specialist only when the title is obviously one's domain (`specialist_key_for_area`: the default departments' slugs and titles, plus Sales) and no department holds that specialist yet; anything else is informational. It refuses a goal the area (or, with no area, the goals with no department) already tracks (pointing at `update_department_goal`), caps a turn at 5 new goals (a sweep is not a stated goal) and an area, or the goals with no department, at 8 (common practice keeps 3–5 key results per objective, and every goal renders in the 4000-character org block, so one area cannot crowd the others out), returns only the exception type on a store failure, is audited with its rationale, and is never offered to an unattended run (`schedule_tools.UNATTENDED_WITHHELD_TOOLS`: reflection, research and the scheduler's proactive trigger). The solo persona teaches it; team mode learns it from the tool description, so the team persona stays byte-identical.\n\nA department can also list **watched entities** (`watched_entities`, edited one per line on its page): vendors it is evaluating, competitors in its lane, tickers it tracks. The research watch policy treats them as strong grounding — a proposal about one can be added to the watch list on its own — and routes the resulting watch, its alerts and any suggestion cards to the department and its head rather than the principal (see External monitoring).\n\n### Solo and team mode\nAn install is either a **team** (the default) or **solo** — one person using Open Executive just for themselves. The mode lives with the user's time zone in its own one-row table, `workspace_settings` (`openexecutive.memory.workspace_settings`, `GET`/`PUT /workspace`), not on the company profile, because onboarding's commit and the form wizard rebuild the profile from scratch and would wipe it. Changing it is the principal's call (open while no principal exists, so first-run setup can choose).\n\nThe same row holds the **principal's role** (`PrincipalRole`, every field nullable): `role_kind` (`owner`, `in_house`, `independent` or `other`), `role_title`, `reports_to` (free text, e.g. \"VP of Operations, Dana Ruiz\"), `remit` (what they are responsible for) and `measured_on` (what they are judged on). Solo is about who uses Open Executive, not the principal's level — a business owner, an executive inside a larger organisation, an independent or fractional executive — and this is the context the solo persona reads the role from. Only solo mode reads it; a mode switch keeps it. `GET /workspace` returns it only to the principal (nulls for anyone else). The columns are added to an existing table additively, and a read of an older table (no role columns yet) never migrates it and reads the role as unset.\n\nSolo also changes how the Executive speaks and what it can do. It works for one person — the principal — without assuming their role (an owner, an executive inside a larger organisation, or an independent); the people in their world are contacts it helps them work with, not a team it routes to. The cached org block becomes the principal's own line (\"Your Principal\") — followed by their role lines when a role is set (a fixed lead-in marking the role private to the principal — use it to tailor advice and drafts, share it with no one else — then `- Role: <title> — <kind in plain words>`, `- Reports to:`, `- Responsible for:`, `- Measured on:`, each only when set, sanitized and capped) — plus their goals grouped by **area** (a department row is an area; no authority levels, heads, cadences, channels or roster), and posting to a department room, broadcasting to the company and assigning a department head are not offered. Specialists never see that block, so each consult in a solo turn also carries a `<principal_role>` tag in its user turn (kind, title and remit). The role is pinned on the session for the turn with the mode (`pin_turn_principal_role`), so the org block, the tag and any workflow the turn starts agree even if the role is edited mid-turn, so a VP inside a large company is told to make the case to their CFO rather than to raise a round. A follow-up the Executive schedules to the principal goes straight to them — its department and scope are dropped, so the authority gate never files a card asking the principal to approve a message to themselves. Meeting bookings are gated on the `meeting_scheduling` class mode, not a department (see below).\n\nSolo turns the department rhythm off without touching the org: switching to solo cancels every pending `dept_cadence` row, `bootstrap_cadences` does nothing (the check sits inside it, so boot, the fixture reset and client-slot activation all honour it), and the scheduler cancels any stray check-in without running or chaining it, and a check-in already running across the switch schedules no successor. Department and people rows are kept, so switching back to team re-schedules the check-ins at once. The boot-time org-completeness warnings (departments with no head) are skipped in solo. A chat `Session` can carry an explicit `workspace_mode` override (read through `effective_workspace_mode`), so a caller that runs many conversations on one Executive — evals, for one — can pick the mode per session instead of flipping the install-wide setting; a `principal_role` override (read through `effective_principal_role`: the override, then the turn's pin, then the setting) does the same for the role.\n\nIn the UI, a `WorkspaceProvider` in the root layout reads the mode and role once (and again after Settings changes them) and reads **team** while loading or on error, so a team install never flashes the solo UI. Solo swaps the sidebar's Company group for **You** — Goals, People and the profile. `/goals` reads `GET /goals` and groups every goal by owner (the viewer first), department (shown once one exists; \"area\" in solo) or status, and adds, edits, moves and deletes them through the department-free goal routes. Team mode lists Goals in the Company group too, and shows the **Departments** tab only once a department exists (`GET /workspace`'s `has_departments`; the Departments pages keep it, since that is where the first is added). Adding a goal asks only for the goal itself: the timeframe is a one-click chip (this quarter by default, filled with the current period's label), target and current are optional, the owner defaults to the viewer, the department to none, and a new goal starts `on_track` until its check-in or the weekly goal review grades it. Adding a teammate on the People page (or moving a contact onto the team) while solo switches the workspace to team mode on its own and says so, with a link to Workspace settings to switch back. The profile entry follows the principal's role kind: **Business profile** for an owner, **Your work** for anyone else, including an unset role or one `GET /workspace` hides from a non-principal. One rule (`profileWording(mode, roleKind)` in `navConfig.ts`) also words the `/company-profile` breadcrumb, intro and section headings, the form wizard and the draft review; it is copy only, so no profile field is renamed, and team wording is unchanged. `/departments` and `/people` stay reachable by URL.\n\nThe setting is per company: a blank client slot starts on the defaults, a saved slot keeps its own (restoring one runs the additive column migration), a fixture load resets it and applies the fixture's optional `workspace.yaml` (`mode`, `timezone` and the optional role keys; read and validated before anything is swapped; a bad value is skipped, never a failed load; the load and unload responses report the mode and zone only, never the role), and the factory reset returns it to team with no zone and no role. The user's own settings survive a demo: the backup a fixture load takes (or reuses) always gets the live settings written to its `workspace.yaml` while the live state is the user's, and an unload from an older backup without one leaves the settings as they are. A zone that cannot be loaded is never fatal — the API answers 422, and a stored one reads as unset.\n\n### Authority gate\nEach department declares an `AuthorityLevel`, which `departments.authority.gate_action` checks whenever a department-scoped scheduled action comes due. The check-in cadence is the exception: it sends nothing itself, and each action it proposes is run through the gate and shown with its verdict in the report — so a propose-only department's check-ins actually run instead of becoming a one-shot approval card. `auto_execute` (autonomous) fires directly. `propose_only` turns the action into an approval card for an approver (`propose_via_alert`) and sends nothing, deferring the card to the approver's next availability window when they are outside it. `escalate` also sends nothing until someone approves: it raises an urgent `[ESCALATION]` card (severity high) right away, never deferred — filed unrouted when nobody holds the scope and no principal is set, since the Briefing lists every unread card — what the Departments page promises (\"the specialist will not act — it forwards everything to a human\"). It used to raise the card *and* dispatch, so approving the card sent the action a second time. The scheduler records which action each card holds (`alert_held_actions`, written only as it files the card), so approving one (`POST /alerts/{id}/approve`, see Today) queues that exact action again without its department, the approval being the consent the gate waited for, and records who approved it. Calendar bookings follow the same rule in team mode. **A team member's chat turn goes through the gate too** (`orchestrator.department_gate`). On a turn whose caller is a team member other than the principal (attended, not private to the principal), every send or edit is checked before it runs: Take the lead's acting tools less `assign_open_loop` and `run_workflow` (`GATED_TOOLS`), and every connected tool `take_the_lead.mcp_tool_acts` flags, the Gmail send or a Docs edit among them (that check fails closed: only a name that starts with a read verb passes). The level is the strictest of the speaker's departments; no department reads as `propose_only`. `auto_execute` runs it. `propose_only` runs it when the speaker could approve it themselves (they hold a People approval range for every kind it touches, money, a contract, a people decision or the board, or wildcard), else holds it. `escalate` always holds it, for someone other than the speaker. A held action is a `department_action` decision with a card on its approver's Today (high severity when escalated; the principal's when it reaches someone outside the team), audited `department_action_held` with the speaker as actor; the model is told it waits and not to retry. Its card shows every field of the call in full and only then offers Approve, which makes that exact call on the server. Under `escalate`, putting someone on the chase and starting a workflow are refused, since they lead to sends later and can't wait on a card. Someone signed in on the web but on no People entry, or a chat-channel sender matched to no one (in practice Google Chat, the one channel that does not turn unmatched senders away), is gated like someone in no department. Google Chat matches its sender by the Google account email on its signed event and hands that entry to the gate alone (`Session.gate_sender_person_id`, never `caller_person_id`), so the principal's own sends there go and a teammate gets their department's authority; email is left to the mailbox's own gates. A gate that can't be read holds. Before this only unattended work and `propose_meeting` met the gate, so a propose-only or escalate department held nothing its own people asked for in chat. The principal's own turns are not gated, nor are unattended runs (Take the lead's gate and this section's scheduler gate cover those). In solo mode there is no operations department to consult: the booking is decided by the `meeting_scheduling` class mode — `propose` puts it on the briefing for whoever holds the **Meetings** approval range on People (`meeting_scheduling`), else the principal — a booking with a contact on it always stays the principal's — and `auto_execute` books right away only when the principal started the turn on a surface that verified it is them (the roster-write rule: the web chat, their own Slack or Discord, or a private Telegram chat with the webhook secret set). A booking asked for in an inbound email, by a contact, or by an unattended run is proposed to the principal instead, so text someone else wrote cannot put a meeting on their calendar. The principal sets it with `GET`/`PUT /decisions/classes/meeting_scheduling` (`{\"decision_class\": \"meeting_scheduling\", \"mode\": \"propose\" | \"auto_execute\"}`; the PUT is principal-only and audited). The approver is chosen by `AuthorityScope` — a scheduled action's `required_scope` (e.g. `vendor_onboarding`, `spend_lt_10k`) feeds `find_approvers(scope)`, which ranks non-principal scope-holders ahead of `WILDCARD` holders. A null/unknown scope falls back to `WILDCARD` (the principal).\n\n### People & channels\nOne rule (`people/reach.py` `reach_order`) orders the channels the Executive reaches a person on, for briefs (`runner.delivery_order`), messages (`message_person` and the follow-up channel pick) and nudges (`prefer_channel_for()`, which also honours availability windows): their `preferred_channel`; with `any`, the chat app they last talked in (`last_channel`, written by `note_talked` at most hourly on a Slack, Discord or private Telegram turn from their own stored id); then Slack → Discord → Telegram; then email as the backup (first for an `email` preference). Each caller still skips a channel that isn't set up or an id it can't use. Departments can also set `slack_channel_id` / `discord_channel_id` / `telegram_chat_id` to broadcast to a team room instead of DMing the head.\n\nEach person's page also lists their **open loops** (see Attunement) — what they committed to or were asked for in conversation — with a *Mark done* action; archiving a person closes their open loops, since an archived person can no longer be chased. The list is the principal's or that person's only; `people.store.is_principal_or_self` is the one check behind it, the close action and reply feedback.\n\nThe Executive can change the roster from a chat turn (`upsert_person`, `archive_person`, `set_department_head`) only when the principal asks, from the web chat, their own Slack or Discord, or a private Telegram chat when `TELEGRAM_WEBHOOK_SECRET` is set to a value Telegram can send. A teammate, an email (even one from the principal's address), Google Chat, the CLI, the MCP server and unattended runs are refused, because a team member's People row is also the web sign-in allow-list, the outbound-email allow-list and approval routing. The People page follows the same rule: adding, editing and archiving people (`POST`/`PATCH /people`, `POST /people/{id}/archive`) are the principal's alone, or anyone's while no one is principal yet, so a first setup can add its owner (`chat._caller_is_principal_or_unclaimed`). Anyone else gets 403. That includes a teammate editing their own entry: its email and chat ids are how they sign in and how the Executive reaches them, and the rest decides their approvals and how they are chased. Their working style and open loops stay theirs. A signed-in owner whose email is on no People entry is refused too; they sign in with the email on their entry, or re-run the setup interview, which fills a missing one with the email they signed in with. The one roster change that can come from an email is the principal's tokened reply to a roster-request email, and the one a chat turn can make beyond these tools is `resolve_roster_request` — both below, both the principal's alone.\n\n### Each person's notifications\n\nEvery team member sets their own on Settings → Notifications (`GET`/`PUT /notifications/settings`, see API), stored one row per person in `person_notification_settings` (`people.notifications`): **quiet hours** (off by default; from–to in their zone, which may cross midnight; every day, or weekdays with the weekend quiet all day), **delivery** — *As they happen* (the default) or *In my briefs* — and **Let urgent things through** (on by default). No row means the defaults, which hold nothing back.\n\nThey apply only to the Executive's **own** outreach: a DM sent while a proactive source is tagged (`attunement.outcomes.tag_proactive` — the scheduler's nudge dispatch, the reflection, research findings, alert review). Replies in a conversation, a message someone asked for in chat, a follow-up someone scheduled (it was asked for at a time), the briefs themselves and reminders a person set are never held. Quiet hours are deliberately not part of the outbound guard's availability check, which runs on every DM — the brief included — and would refuse a brief scheduled inside the person's own quiet hours.\n\nTwo places apply them. **Nudge routing** asks `people.channel` with `quiet_hours=True` (`is_within_availability`, `next_window_for`, `prefer_channel_for`, `next_available_window`), so the nudge engine and the scheduler's `proactive_nudge` re-check defer a nudge to the end of quiet hours exactly as they defer to an availability window — nothing is written at 3am. **Every other proactive DM** meets `outbound_guard.check_hold` in the send handlers once the anti-spam guard has allowed it: if the recipient is in quiet hours, or chose *In my briefs*, the text is stored in `held_messages` (`people.held_messages`) and the tool returns `{\"status\": \"held\"}`; `message_person` stops there instead of trying another channel; alert review counts a route as made but not a nudge toward its cap, and holds no second nudge about an alert while one waits. If it can't be stored it is sent. A scheduler dispatch whose message was held is audited as `held`, not `delivered`. Quiet-hours rows go out when quiet hours end, as one *While you were away* message (see Scheduler); brief rows ride the principal's next morning brief or end-of-day digest as one *While you were away* block, and anyone else's go out on their own when those briefs run, since only the principal has a brief. Switching back to *As they happen* releases what waited. Whoever sends a row claims it first, so a brief and a release pass never both send it. Someone on leave or outside their availability windows is postponed to when that ends; any other failed delivery backs off (2, 4, 8 … minutes, at most an hour). Rows are capped at 50 per person and dropped after 7 days undelivered, each drop audited privately to the person as a count; a brief row with no brief to carry it for 36 hours goes on its own.\n\n**Urgent** messages — approval nudges (`nudge_stalled`) and alert-review DMs about a critical (`urgent`-severity) alert (`notifications.urgent_send`) — get through when the person lets urgent things through. A workflow approval question (`gate_delivery`, `notifications.approval_send`) is never held, whatever the setting, since only a reply to it resolves the gate. If quiet hours cover every availability window, a nudge falls back to the next window rather than being dropped. Not covered: nudges on the email channel (the MCP mail tool bypasses the send handlers, so routing defers them for quiet hours but they are never held for a brief), department channel broadcasts, and the alert dispatcher's own channels (the web stream, a self-addressed email, and a Slack DM that is still a stub). Department check-ins send no DM: what they propose becomes approval cards.\n\n### Team and contacts\nA `Person` is either on the **team** (`kind = \"team\"`, the default) or a **contact** (`kind = \"contact\"`): a client, contractor or advisor the Executive may need to email or invite but who must not get what a team row grants. Before contacts existed, emailing a client meant adding them as a Person, which also let them sign in and talk to the bot.\n\nContacts are **denied by default**. `people.store.list_people`, every `find_person_by_*` (and `find_person_by_channel_ref`), and the `people.registry` cache return team members only unless the caller passes `include_contacts=True`; `find_approvers` has no opt-in at all. So a contact is never in the web sign-in list (`GET /auth/allowed-emails`), the Slack / Telegram / Discord inbound gates, approvers, the alert review's roster slice, `/today`'s people list and insights, open-loop owners (an ask of a contact opens no loop, so the nudge engine never chases them), a workflow gate's approver or cadence recipient, the recap DM after a meeting, or a department head (`set_department_head` and `PATCH /departments/{slug}` refuse one). `get_person(id)` returns either kind and its security-relevant callers check `kind`.\n\nThe roster reads that **opt in** are few: the Executive's egress checks (Gmail, Calendar and Drive in the MCP gateway, calendar attendee resolution, and the chat DM tools), the People page's listing (`GET /people?include_contacts=true`), the chat `list_people` tool (both kinds, with `kind`), the org prompt block, and the roster-request intake (which looks a sender up with contacts included only to leave a contact out). Even there, reaching a contact needs **the principal's own turn**: `people_tools.contacts_reachable_now()` is true only on a turn the principal started on a surface that verified them — the same rule as roster writes above — or inside `grant_contact_egress()`, which the decisions approve route uses when the approver resolves to the principal. An inbound email (even from the principal's address), a teammate's turn, and unattended runs (scheduler, workflows, alert review) can still reach the team but not contacts. A workflow resumed by `apply_resolution`'s kick runs with no chat session, like the resumer's poll loop, so it never inherits the approver's turn.\n\n**Contacts are private to the principal.** A solo user may be an executive whose contacts include their boss, reports, board and clients, so nobody else using the install sees one. The rule is `contacts_reachable_now()` again: off the principal's own verified turn a contact does not exist — `list_people` returns the team with the same count, a contact's id or address gets exactly the error an unknown one gets (never a contact-specific refusal), and the org prompt block has no Contacts section. The API applies the same rule by caller (`people.caller_is_principal`: the signed-in email is the principal's, or no `x-caller-email` — CLI, local login; unlike the owner rule above, nobody is while no one is principal yet): `?include_contacts=true` is ignored for anyone else, a contact's id is a 404 on `GET /people/{id}` and its open loops exactly like a missing one (editing and archiving are already a 403 before any lookup), only the principal may create a contact or turn someone into one (403, a first setup with no principal included), and `GET /people/me` tells the UI whether to offer contacts at all.\n\nWhatever the Executive raises about a contact stays with the principal too. A meeting with a contact is proposed to the principal (not the scope's approver), its decision is hidden from everyone else's `/decisions` (404 by id), and its card carries `alerts.models.PRIVATE_ALERT_TAG`. So does an alert raised on a turn about the principal's private mail (`Session.private_to_principal`, set by the email poller for mail from a contact and for mail the principal forwards): the pipeline routes it to the principal, stores it without a live push or broadcast, and dedups it only against other private alerts. Private alerts appear on the principal's own `/today` and chat digest only — never a teammate's, the activity rail, the narrative, the morning brief, the end-of-day digest, the reflection or the alert review — and a teammate's ack, reopen or bulk sweep cannot touch one (404). A private turn itself reaches the principal and nobody else (email, DMs, invites). It is not offered, and cannot run, the tools that post to other people, publish where the team reads (artifacts, goal updates, saved workflows, skill drafts) or start work outside the turn (`schedule_tools.PRIVATE_TURN_WITHHELD_TOOLS`), and of the MCP tools it may use only Google Workspace reads and the Gmail send the gateway narrows to the principal. Its audit rows are the principal's alone, as is any row on the principal's own turn that names one of their contacts (see Audit Log). DMs to a contact stay off the activity rail.\n\n`upsert_person` takes `kind`: explicit wins; an update that omits it keeps the row's kind (the store's full-row UPDATE preserves it, so nothing promotes a contact to the team by accident); a new person defaults to `team`, or to `contact` in solo mode. On the principal's own verified turn the org prompt block lists contacts in a short `## Contacts` section after the team block (or, in solo, after `## Your Principal` and their goals) — name, role/company and whether an email is on file, capped at 25 — in both modes. Every other turn gets the team block, byte-identical to an install with no contacts, so the 5-minute cached block has two stable variants.\n\nOn the **People page** the principal sees **Team** and **Contacts** tabs; anyone else sees the team only (no Contacts tab, no kind selector). Solo mode opens on Contacts, and its Team tab shows only the principal. The add form has a kind selector and hides authority scopes, reply SLA and the principal checkbox for a contact; a contact's page hides scopes, SLA, availability windows, open loops, outreach and working style. Adding a team member in solo offers to switch the workspace to team mode, or keep solo.\n\n### Contacts the Executive keeps up to date\n`people.contact_keeper` keeps the principal's contacts current on its own, from **their own mailbox** under Act as me (`delegation.gmail_for`: Gmail or Outlook, never the MCP gateway). It runs in the background (the scheduler starts it every 6 hours), never in a chat turn, and only while the principal has Act as me on with their own mailbox connected. The switch **Keep my contacts up to date** (`contact_keeper_state.mode`: `on`, the default; `ask`; `off`) sits on the Contacts tab.\n\n- **Adds**: someone the principal emailed at least twice in 60 days (sent mail they wrote: no lists, bulk or auto mail, nothing Handle it for me sent, at most 10 recipients) or met (meetings from 30 days back to 14 ahead that they organised or accepted, at most 10 guests). Names come only from To/Cc headers and guest lists, never a mail's text. Skipped: their own and the Executive's addresses, anyone in People (archived included), role and no-reply addresses, company-domain addresses (`identity.is_company_address`), and anything blocked. At most 10 per run.\n- **Updates**: an alias when a known contact (two-word full name) is mailed twice at a new address under that exact name as the principal's To line wrote it, always as a suggestion whatever the switch (mail to the contact may go there, so one reply to a lookalike must not attach it; an undone or archived address never is); a fuller name for a one-word contact; a role (\"Head of Procurement, Acme\") read from the contact's newest mail (`latest_from`) when its sender is authenticated and the From address is theirs. The role read is one forced-tool model call on the mail's last 1,200 characters, kept only if every word is in the mail, it is at most 10 words and 80 characters, and it carries no link, address or instruction. At most 5 role reads per run and 20 a day, and an address again only after 14 days. A field the principal typed is never overwritten: a name or role is written only when empty or last set by the keeper.\n- **Record and Undo** (`people.contact_changes`): every change is a row (`contact_changes`) with its reason and evidence. Under `ask` it is stored as a suggestion and applied on **Add**. **Undo** archives an added contact or restores the old value if nobody changed it since; Undo and **Not now** block that address, field and value (`contact_keeper_blocks`) so they are never proposed again. A contact added from chat is recorded too (reason `asked`), so it has an Undo. Each run writes one private `contacts_kept` audit row with counts, no names.\n- **Contact details**: the same signature read copies the sender's other contact details (`Person.details`, table `person_details`: phone, mobile, website, LinkedIn, another profile, postal address). Each is kept only when the mail writes it (a phone's digits as one number, every word of an address, a link's host and path), a website only on the sender's own mail domain (never a free mail one) and a profile only on a known site (`people.contact_info`). One of each kind is the keeper's: a newer one replaces it, and a kind anyone typed a value for is left alone (no second phone next to a typed one). Each is a change with **Undo**. People and chat (`upsert_person`'s `details`) can type them too. The people routes and `list_people` return a person's details only to the principal and to that person; everyone else gets the row with `details` empty. They are for the reader only: nothing is ever sent to a phone or link.\n- **Team members** get the same signature read (role when empty, contact details) and the same note; their name and addresses are never changed.\n- **What it remembers** (`people.contact_notes`, table `contact_notes`): for every contact and team member, when the principal last emailed, heard from (the date of their newest authenticated mail) and met them, their next meeting, counts, and a short note with up to 6 topics on what they deal with each other about. The note is written by one forced-tool call only from the principal's **own** sent mail to that contact (their new text, without quoted mail; nothing Handle it for me sent) and the titles of meetings they organised, so nobody else's words go in. It is rewritten only when that mail or those meetings change, at most 5 a pass and 20 a day, and a note holding a link, address or tag is dropped. \"Last heard from\" is read for up to 20 people a pass, whatever the note budget. The Executive reads it on the principal's own turn in `list_people` (contacts and teammates) and `ask_about_person` (contacts), to suggest who to ask; the person's page shows it to the principal as **What I remember**. A teammate's note is the principal's alone too: it comes from the principal's mailbox, never the teammate's own mail to the Executive or their peer memory. Archiving someone deletes it.\n\nAll of it is the principal's alone: the routes answer only `caller_is_principal`, and the rows are private. An added contact is an ordinary contact, with the rules above.\n\n### Addresses: aliases and company domains\nA `Person` has one **primary** address (`email`) — the one they sign in with — and up to 10 **aliases** (`email_aliases`, the `person_emails` table: `email_norm` unique across the roster, `source` `manual` / `chat_tool` / `roster_request`, deleted with the person or when they are archived, so an archived person never blocks the address). An alias matches their mail and may be emailed, but **is never a login**: `find_person_by_email` stays primary-only and is what web sign-in and `/auth/allowed-emails` read; mail goes through `find_person_by_address` (primary or alias, case-insensitive, the team row winning over a contact) and `people.identity.resolve_email_sender`. Aliases are edited on the person's page (\"Other addresses\") and by `upsert_person`'s `email_aliases` (the full list, replacing the current one); an address that is already someone else's primary or alias is refused (`AddressInUseError`, a 409 on the API), and so is a new primary that is someone's alias. Onboarding still imports no addresses beyond the owner's own sign-in email, and an alias counts as taken there too.\n\nOn top of exact matches, `resolve_email_sender` applies a **company-domain rule, team only**: on one of the company's own domains an address's local part (lowercased, `+tag` dropped) matches a teammate whose own address on any company domain has that local part — `[email protected]` is the Anna whose address is `[email protected]`. A domain alone proves nothing: an unknown local part on the company domain matches nobody (it becomes a roster request, pre-filled as a teammate), and when two teammates would match, nobody does. The company's domains are the `company_domains` workspace setting (Settings → Workspace, `PUT /workspace`, at most 10), or, when that is unset, the domains of the principal's own addresses — never a free-mail provider's (`identity.FREEMAIL_DOMAINS`: gmail.com, outlook.com and the like, refused as a setting too), where a local part says nothing about who someone is. The outbound gates build `RosterAllow` from the same rule, so the Executive may email exactly the addresses inbound mail would match (see Integrations).\n\n### New senders: roster requests\nWhen someone not on the roster — and not one of the principal's contacts — writes in (an email address, a Slack or Discord user, a private verified Telegram chat), their message is held as a **roster request** (`people.roster_requests`: tables `roster_requests`, `roster_request_messages`, `roster_ack_log`) instead of being dropped; they get a fixed acknowledgement (an email sender only when Gmail authenticated them — see Integrations) and the principal gets a private card on `/today`, which says whether they were actually told (`surface_card(acknowledged=...)`, from `ack_sent_at` after the acknowledgement is attempted) and, for a chat or company-domain sender, a prompt on their delivery channel (the per-channel mechanics are in Integrations). A request holds at most 10 messages (later ones only count), is one per `(channel, sender)` while pending, and suggests a match only from a chat account whose profile email is already someone's (never from a display name, which the sender chooses) — and `suggested_kind = team` for an address (or a chat profile email) on a company domain.\n\nThe principal answers one of three ways, all ending in `roster_requests.resolve` (a claim on the row, so two answers race and one wins; nothing is written unless it succeeds):\n\n- **approve** — add them as a new Person with this address or chat id, `team` or `contact`;\n- **link** — they are someone already on the list writing from somewhere new: an email becomes that person's alias — never their primary, even if they had none, so a link never creates a login, a chat id fills that person's field (replacing a different one only when the principal says so; otherwise `ChannelIdConflict`);\n- **decline** — leave them off; they are not raised again for 30 days.\n\nOnly the principal answers, and only where they are verified: the **web card** (`POST /people/requests/{id}/approve|decline`, which requires an explicit team/contact choice), their **own Slack or Discord DM or verified private Telegram chat** (the `resolve_roster_request` chat tool, which answers only ids the server showed that turn in its `<roster_requests>` block — see Agents), or an **email reply from their own primary address** (not an alias) carrying the request's one-time token, which the gateway hides from every other read of the Executive's mailbox so no model turn can copy it. When the chat or email answer does not say team or contact, the default is team for a sender on a company domain, else contact — the least access. Whoever answers, the held messages are then replayed as if the sender had been on the roster all along (a new contact gets only held email, on the private contact path); a decline drops them. A roster write outside a request — the People page (`POST`/`PATCH /people`) or `upsert_person` — reconciles too: any pending request whose sender now matches someone closes as `superseded` and replays.\n\nUnanswered requests expire after `ROSTER_REQUEST_TTL_DAYS` (14). A held message's text is kept only until the request is resolved or expires (then purged, even on decline), its card, routes and audit rows are private to the principal (404 for anyone else), and the tables are wiped with the rest of the People data on a company swap or fixture load.\n",+ "markdown": "The org model has two persistent registries backed by SQLite (both share `episodic_memory.db`): **departments** and **people**. A new install starts with **no departments**: nothing is created at boot. Setup's commit creates only the departments its draft names (none for a solo workspace or a team that names none), and the Executive's `create_goal` tool creates one when a goal names an area that has none. A department named after one of the eight catalogue areas in `departments/charters` (Strategy, Finance, People & Talent, Legal, Operations, Marketing, Product, Board & Investor Comms) starts from that area's charter, and setup and `create_goal` wire it to the matching specialist (at most one department per specialist). Workspaces created before this keep the departments they have. `seed_default_departments` still inserts the full catalogue once per database (tracked by a `default_departments_seeded` sentinel in `departments_meta`), but only tests and tooling call it.\n\n### Company setup (onboarding)\nBoth registries are first populated by **conversational onboarding** at\n`/onboard`. The user describes their business in their own words (optionally\nattaching a deck or one-pager); the `onboarding_interviewer` agent — Council-\nvisible but outside `SPECIALIST_REGISTRY`, so the Executive cannot reach it via\n`consult_specialist` — asks at most four quick questions, each with two to four\nsuggested answers the UI shows as buttons (`Question.options`, kept per session\nlike the hint), and never about departments or finances. It then emits a draft\ncompany profile, the people, this year's goals (`GoalDraft`: a title, an\noptional target, an owner by name, blank meaning the principal) and departments\nonly when the user's own words already listed people by department. The user\nreviews and edits that draft inline (the same section editors `/company-profile`\nuses, with empty optional sections left out) before anything is written; its\nnotes show as \"Please check\" in plain sentences, and \"After you save, I'll\"\nlists the research and weekly goal checks that follow. Until a profile exists\nthe owner (the principal) is sent to `/onboard` from Home and every shell page\nexcept Settings, unless they chose \"I'll do this later\"; Save goes straight to\nHome, whose banner slot carries a getting-started checklist (profile, first question,\nthe two mailboxes, the team) until every required row is done or it is hidden\n(`ui/src/lib/firstRun.ts`).\n\nTwo properties matter here. First, **the interview never writes the profile**:\nthe exchange lives in a TTL'd session (2 h, 50 at most) that the API process\nholds and also writes through to the `onboarding_interviews` table of the\nepisodic SQLite database after every turn (`onboarding/interview_store.py`),\nso a restart or another worker finds it again (a newer stored row replaces\na held copy, so two workers never overwrite each other) and the browser,\nwhich keeps only the session id, resumes at the same question or draft via\n`GET /onboard/interview/{id}`. A session answers only to the caller who\nstarted it; anyone else gets the 404 an unknown id gets. Every answer stays on screen with an Edit link, on the draft review too: `POST /onboard/interview/revise` replaces one earlier answer (an edited description keeps its attached files), drops everything after it, resets the question count to what was asked before it, and runs the interviewer again, so it asks again from there or redrafts. A failed turn leaves the session as it was. Start over (on the \"Picking up where you left off\" banner) deletes the session at once with `DELETE /onboard/interview/{id}`. The single profile write happens at\n`POST /onboard/interview/commit`, so a rejected save leaves the draft editable\nand retryable, and a successful one deletes the stored transcript at once (it\ncarries the company's financials). The commit then saves the goals\n(`onboarding.commit.save_onboarding_goals`): for this year, with no department\nunless they name a drafted one, owned by the named person or else explicitly by\nthe principal, and additive (a goal worded like one already set for this year, or ongoing, is skipped, so a\nre-run adds no duplicates). Their titles also become the profile's\n`strategic_priorities.current_year`. Second, **the department step is additive** — drafted\ndepartments are matched onto existing ones by slug or title and updated in\nplace, unmatched ones are created, and nothing is ever deleted. That is\ndeliberately unlike the fixture loader's `_seed_departments`, which DELETEs\nevery row first: doing that here would drop departments the workspace already\nhas, with their goals and `specialist_key` wiring, on a re-run of setup. A\ndrafted department that matches nothing is created, wired to the specialist\nwhose area it is named after (`specialist_key_for_area`) unless another\ndepartment already has that specialist. Matching goes through\n`departments.store.match_department` (slug as given or as the store would slug\nit, else case-insensitive title), the same rule the `create_goal` chat tool uses. `profile.org_structure` is derived from the reviewed\npeople and department tables at commit time, and this year's priorities from\nthe goals, so neither can drift from them.\n\nExactly one person is marked `is_principal` and receives `WILDCARD` authority;\nno contact details are imported, even if the user volunteers them, because\nchannels and handles are a deliberate action on the People page. The one\nexception is the owner's own sign-in email: the review screen has a separate\nfield for it, pre-filled with the current owner's email on a re-run and with the\nsigned-in user's on a first setup. The commit checks it before writing anything\n(malformed, already another person's — as their address or one of their\naliases — or different from the owner's existing\nemail is a 422 — setup only fills in a missing one) and then saves it on the\nprincipal's row. One case is adopted rather than rejected: when the address is\nthe one the caller signed in with and sits on the current principal's row\nunder another name (a hosted sign-up seeds that row from the address when no\nname was given), `adopt_owner_row` renames that row to the drafted name\nbefore the roster upsert, so the first save updates the owner's entry instead\nof failing with \"use the owner's name exactly as on the People page\". Without it the signed-in owner matched no Person,\nso their chat history stayed empty until they found the People page. With\nlocal login (`make dev` without Google or SSO sign-in) the field is optional: that\nsession carries no email, which the backend already reads as the principal.\nBecause the save demotes the old principal, only the current owner may draft\nsomeone else as \"This is me\" when setup is re-run (403 otherwise); a re-run\nthat keeps the owner, or a first setup, is open to anyone who can sign in. On such a re-run, someone who isn't the owner may fill in the owner's missing email only with the address they signed in with (403 otherwise): the roster is the sign-in allow-list, so any other address would let them sign in as the owner. That keeps the way in for an owner whose entry has no email, which is also where Setup status and the Executive send them, since the People page refuses anyone it can't recognise as the owner. Re-running setup stays open to everyone deliberately: the Company Profile and Departments pages already let anyone signed in change the same things. The original\n12-step form wizard remains available at `/onboard?mode=form` and backs the\n`openexecutive onboard` CLI. It adds people with their emails, so once there is an owner only they can save it.\n\nA first setup opens on one free-text box: the user describes their work in their own words, optionally attaching a one-pager. `POST /onboard/interview/understand` reads that text once (no session, nothing stored) and reports what it already says: Personal or Team, the principal's role (owner / founder, an executive inside an organisation, independent or fractional, or other), title, who they report to, the company and this year's focus. Nothing is asked about them: the UI writes `PUT /workspace` straight away (the mode read, the browser's time zone and, for Personal, only the role fields the text gave) and starts the interview with the same text. The mode is Team when the text names colleagues at the user's own company, Personal when it is only the user or the only people named are outside the team (a boss, clients, partners), and Personal when unclear or when the read fails. The review shows the choice on one line (\"Set up for your team: …\") with a Change link, saved with the draft. A re-run over an existing profile skips the box and keeps the saved mode. The tone of voice is set in Settings. In solo the interviewer's first user turn carries a short\nhint, one static constant per role kind (`onboarding.interview.solo_hint`):\nalways draft only the principal, as principal, no departments, and every goal\ntheirs; an owner is asked about the offer, customers and pricing; an in-house\nexecutive about their organisation (the company profile is their employer),\ntheir function and remit, their manager, their team size as context rather\nthan a roster, and what they are measured on; an independent about their\npractice, clients and pricing; with no role given, the role-neutral hint asks\nwhich of these they are. Each also asks their top goals. The hint never\ncarries the principal's own text, and it rides in the user turn, so the cached\nsystem prompt is the same constant in both modes. The solo review screen shows only **You**\n(name and role) and the sign-in email, and commits that one principal with no\ndepartments — the existing departments stay, as the areas goals are grouped\nby. The form wizard (and the CLI) read the mode when a session starts and skip\nthe Team Size, Team Members and Fractional Executives steps in solo; each\nstatus response numbers the steps it actually asks and flags whether the\ncurrent one is `optional`.\n\n### Departments\n**Goals stand on their own.** A `Goal` has an owner and, optionally, a department (`department_goals.department_slug` is nullable; `owner_person_id` is nullable too). Its owner is resolved at read time (`departments.goals.effective_owner_id`): its own `owner_person_id`, else its department's head, else the principal, so a goal that follows its department moves with a new head. A goal no department check-in grades — one with no department, or in a department with no specialist — is graded weekly by the `goal_review` scheduled kind (`departments.goal_review`, Monday 09:00 in the user's zone, both modes) with `cso` (`goals.FALLBACK_REVIEWER`), through the check-in's own verdict contract and `persist_goal_verdicts`; it skips goals graded within six days and not edited since, so a quiet week makes no model call. Deleting a department keeps its goals: each with no owner of its own takes the department's head as owner, then all lose the department (`store.delete_department`). An old database is migrated in place (`_migrate_goals_stand_alone`: the table is rebuilt once to drop the NOT NULL, then the owner column is added). `GET /goals` lists every goal with `owner_id`, `owner_name` and `department_title`; `PATCH`/`DELETE /goals/{id}` reach any goal and can move it between owners and departments (an owner must be a non-archived team member, never a contact).\n\nEach department carries a charter, `Goal` rows, `ChecklistItem`s, and named cadences. Cadence specs (`daily@HH:MM`, `weekly@DOW@HH:MM`, `quarterly@DD-HH:MM`, UTC) run a `department_check_in` workflow when due (never in solo mode, below), then chain the next occurrence. A new department starts with `check_in: weekly@mon@09:00` (`DEFAULT_CHECK_IN_CADENCE`; it was daily, and departments created before keep theirs). An empty spec turns the check-in off. `PATCH /departments/{slug}` rejects a spec that does not parse with a 422, and a changed `check_in` re-times the department at once (`cadence.reschedule_check_in` cancels the pending check-in and queues the next one on the new schedule, or none when it is off). The department page offers Daily / Weekly (day and time) / Off rather than the raw spec, and shows the next check-in from `next_check_in_at`. The check-in is skipped (and still chained) when the department has no specialist, no Goals, or nothing new since its Goals were last reviewed (a row private to the principal does not count as new: whether it runs is visible to the department). An idle initiative is chased with the department's head, or with the principal when there is none. A department may map to a specialist agent via `specialist_key`, or stay informational. No department is seeded for the `sales` specialist (existing installs keep their departments unchanged), so its consults skip the department-memory prefetch and the Honcho department sync, as `triage`'s do. An assigned head gets a generated department-head voice persona.\n\n**Goals from chat.** Besides the goal routes and the UI, the Executive keeps goals current from a chat turn: `update_department_goal` records reported progress, and `create_goal` starts tracking a new goal the principal states (\"20 paying clients by the end of Q4\"). Neither needs a department: `list_department_goals` without `department_slug` lists every goal with its owner and department, `update_department_goal` names a goal by `goal_id` alone (a `department_slug` it is given must still match), and `create_goal` leaves `area` out unless the principal names one, taking an optional `owner_person_id` (a team member, from `list_people`). Only the principal, on a surface that verified it is them (`people_tools.is_principal_on_verified_surface`, the roster tools' rule: the web chat, their own Slack or Discord, a private Telegram chat with a valid webhook secret), may call it: a goal renders in every turn's org block as one of the principal's own, so a goal from an inbound email, Google Chat, a teammate or an unattended run would be text carrying their authority. Anyone else gets a refusal (audited with the caller and surface) and nothing is written. `create_goal` names the area as free text — a slug or a title, resolved by `match_department` — so its schema is static and adding an area never changes the cached tool prefix. When nothing matches it creates the area (`create_department`), giving it a specialist only when the title is obviously one's domain (`specialist_key_for_area`: the default departments' slugs and titles, plus Sales) and no department holds that specialist yet; anything else is informational. It refuses a goal the area (or, with no area, the goals with no department) already tracks (pointing at `update_department_goal`), caps a turn at 5 new goals (a sweep is not a stated goal) and an area, or the goals with no department, at 8 (common practice keeps 3–5 key results per objective, and every goal renders in the 4000-character org block, so one area cannot crowd the others out), returns only the exception type on a store failure, is audited with its rationale, and is never offered to an unattended run (`schedule_tools.UNATTENDED_WITHHELD_TOOLS`: reflection, research and the scheduler's proactive trigger). The solo persona teaches it; team mode learns it from the tool description, so the team persona stays byte-identical.\n\nA department can also list **watched entities** (`watched_entities`, edited one per line on its page): vendors it is evaluating, competitors in its lane, tickers it tracks. The research watch policy treats them as strong grounding — a proposal about one can be added to the watch list on its own — and routes the resulting watch, its alerts and any suggestion cards to the department and its head rather than the principal (see External monitoring).\n\n### Solo and team mode\nAn install is either a **team** (the default) or **solo** — one person using Open Executive just for themselves. The mode lives with the user's time zone in its own one-row table, `workspace_settings` (`openexecutive.memory.workspace_settings`, `GET`/`PUT /workspace`), not on the company profile, because onboarding's commit and the form wizard rebuild the profile from scratch and would wipe it. Changing it is the principal's call (open while no principal exists, so first-run setup can choose).\n\nThe same row holds the **principal's role** (`PrincipalRole`, every field nullable): `role_kind` (`owner`, `in_house`, `independent` or `other`), `role_title`, `reports_to` (free text, e.g. \"VP of Operations, Dana Ruiz\"), `remit` (what they are responsible for) and `measured_on` (what they are judged on). Solo is about who uses Open Executive, not the principal's level — a business owner, an executive inside a larger organisation, an independent or fractional executive — and this is the context the solo persona reads the role from. Only solo mode reads it; a mode switch keeps it. `GET /workspace` returns it only to the principal (nulls for anyone else). The columns are added to an existing table additively, and a read of an older table (no role columns yet) never migrates it and reads the role as unset.\n\nSolo also changes how the Executive speaks and what it can do. It works for one person — the principal — without assuming their role (an owner, an executive inside a larger organisation, or an independent); the people in their world are contacts it helps them work with, not a team it routes to. The cached org block becomes the principal's own line (\"Your Principal\") — followed by their role lines when a role is set (a fixed lead-in marking the role private to the principal — use it to tailor advice and drafts, share it with no one else — then `- Role: <title> — <kind in plain words>`, `- Reports to:`, `- Responsible for:`, `- Measured on:`, each only when set, sanitized and capped) — plus their goals grouped by **area** (a department row is an area; no authority levels, heads, cadences, channels or roster), and posting to a department room, broadcasting to the company and assigning a department head are not offered. Specialists never see that block, so each consult in a solo turn also carries a `<principal_role>` tag in its user turn (kind, title and remit). The role is pinned on the session for the turn with the mode (`pin_turn_principal_role`), so the org block, the tag and any workflow the turn starts agree even if the role is edited mid-turn, so a VP inside a large company is told to make the case to their CFO rather than to raise a round. A follow-up the Executive schedules to the principal goes straight to them — its department and scope are dropped, so the authority gate never files a card asking the principal to approve a message to themselves. Meeting bookings are gated on the `meeting_scheduling` class mode, not a department (see below).\n\nSolo turns the department rhythm off without touching the org: switching to solo cancels every pending `dept_cadence` row, `bootstrap_cadences` does nothing (the check sits inside it, so boot, the fixture reset and client-slot activation all honour it), and the scheduler cancels any stray check-in without running or chaining it, and a check-in already running across the switch schedules no successor. Department and people rows are kept, so switching back to team re-schedules the check-ins at once. The boot-time org-completeness warnings (departments with no head) are skipped in solo. A chat `Session` can carry an explicit `workspace_mode` override (read through `effective_workspace_mode`), so a caller that runs many conversations on one Executive — evals, for one — can pick the mode per session instead of flipping the install-wide setting; a `principal_role` override (read through `effective_principal_role`: the override, then the turn's pin, then the setting) does the same for the role.\n\nIn the UI, a `WorkspaceProvider` in the root layout reads the mode and role once (and again after Settings changes them) and reads **team** while loading or on error, so a team install never flashes the solo UI. Solo swaps the sidebar's Company group for **You** — Goals, People and the profile. `/goals` reads `GET /goals` and groups every goal by owner (the viewer first), department (shown once one exists; \"area\" in solo) or status, and adds, edits, moves and deletes them through the department-free goal routes. Team mode lists Goals in the Company group too, and shows the **Departments** tab only once a department exists (`GET /workspace`'s `has_departments`; the Departments pages keep it, since that is where the first is added). Adding a goal asks only for the goal itself: the timeframe is a one-click chip (this quarter by default, filled with the current period's label), target and current are optional, the owner defaults to the viewer, the department to none, and a new goal starts `on_track` until its check-in or the weekly goal review grades it. Adding a teammate on the People page (or moving a contact onto the team) while solo switches the workspace to team mode on its own and says so, with a link to Workspace settings to switch back. The profile entry follows the principal's role kind: **Business profile** for an owner, **Your work** for anyone else, including an unset role or one `GET /workspace` hides from a non-principal. One rule (`profileWording(mode, roleKind)` in `navConfig.ts`) also words the `/company-profile` breadcrumb, intro and section headings, the form wizard and the draft review; it is copy only, so no profile field is renamed, and team wording is unchanged. `/departments` and `/people` stay reachable by URL.\n\nThe setting is per company: a blank client slot starts on the defaults, a saved slot keeps its own (restoring one runs the additive column migration), a fixture load resets it and applies the fixture's optional `workspace.yaml` (`mode`, `timezone` and the optional role keys; read and validated before anything is swapped; a bad value is skipped, never a failed load; the load and unload responses report the mode and zone only, never the role), and the factory reset returns it to team with no zone and no role. The user's own settings survive a demo: the backup a fixture load takes (or reuses) always gets the live settings written to its `workspace.yaml` while the live state is the user's, and an unload from an older backup without one leaves the settings as they are. A zone that cannot be loaded is never fatal — the API answers 422, and a stored one reads as unset.\n\n### Authority gate\nEach department declares an `AuthorityLevel`, which `departments.authority.gate_action` checks whenever a department-scoped scheduled action comes due. The check-in cadence is the exception: it sends nothing itself, and each action it proposes is run through the gate and shown with its verdict in the report — so a propose-only department's check-ins actually run instead of becoming a one-shot approval card. `auto_execute` (autonomous) fires directly. `propose_only` turns the action into an approval card for an approver (`propose_via_alert`) and sends nothing, deferring the card to the approver's next availability window when they are outside it. `escalate` also sends nothing until someone approves: it raises an urgent `[ESCALATION]` card (severity high) right away, never deferred — filed unrouted when nobody holds the scope and no principal is set, since the Briefing lists every unread card — what the Departments page promises (\"the specialist will not act — it forwards everything to a human\"). It used to raise the card *and* dispatch, so approving the card sent the action a second time. The scheduler records which action each card holds (`alert_held_actions`, written only as it files the card), so approving one (`POST /alerts/{id}/approve`, see Today) queues that exact action again without its department, the approval being the consent the gate waited for, and records who approved it. Calendar bookings follow the same rule in team mode. **A team member's chat turn goes through the gate too** (`orchestrator.department_gate`). On a turn whose caller is a team member other than the principal (attended, not private to the principal), every send or edit is checked before it runs: Take the lead's acting tools less `assign_open_loop` and `run_workflow` (`GATED_TOOLS`), and every connected tool `take_the_lead.mcp_tool_acts` flags, the Gmail send or a Docs edit among them (that check fails closed: only a name that starts with a read verb passes). The level is the strictest of the speaker's departments; no department reads as `propose_only`. `auto_execute` runs it. `propose_only` runs it when the speaker could approve it themselves (they hold a People approval range for every kind it touches, money, a contract, a people decision or the board, or wildcard), else holds it. `escalate` always holds it, for someone other than the speaker. A held action is a `department_action` decision with a card on its approver's Today (high severity when escalated; the principal's when it reaches someone outside the team), audited `department_action_held` with the speaker as actor; the model is told it waits and not to retry. Its card shows every field of the call in full and only then offers Approve, which makes that exact call on the server. Under `escalate`, putting someone on the chase and starting a workflow are refused, since they lead to sends later and can't wait on a card. Someone signed in on the web but on no People entry, or a chat-channel sender matched to no one (in practice Google Chat, the one channel that does not turn unmatched senders away), is gated like someone in no department. Google Chat matches its sender by the Google account email on its signed event and hands that entry to the gate alone (`Session.gate_sender_person_id`, never `caller_person_id`), so the principal's own sends there go and a teammate gets their department's authority; email is left to the mailbox's own gates. A gate that can't be read holds. Before this only unattended work and `propose_meeting` met the gate, so a propose-only or escalate department held nothing its own people asked for in chat. The principal's own turns are not gated, nor are unattended runs (Take the lead's gate and this section's scheduler gate cover those). In solo mode there is no operations department to consult: the booking is decided by the `meeting_scheduling` class mode — `propose` puts it on the briefing for whoever holds the **Meetings** approval range on People (`meeting_scheduling`), else the principal — a booking with a contact on it always stays the principal's — and `auto_execute` books right away only when the principal started the turn on a surface that verified it is them (the roster-write rule: the web chat, their own Slack or Discord, or a private Telegram chat with the webhook secret set). A booking asked for in an inbound email, by a contact, or by an unattended run is proposed to the principal instead, so text someone else wrote cannot put a meeting on their calendar. The principal sets it with `GET`/`PUT /decisions/classes/meeting_scheduling` (`{\"decision_class\": \"meeting_scheduling\", \"mode\": \"propose\" | \"auto_execute\"}`; the PUT is principal-only and audited). The approver is chosen by `AuthorityScope` — a scheduled action's `required_scope` (e.g. `vendor_onboarding`, `spend_lt_10k`) feeds `find_approvers(scope)`, which ranks non-principal scope-holders ahead of `WILDCARD` holders. A null/unknown scope falls back to `WILDCARD` (the principal).\n\n### People & channels\nOne rule (`people/reach.py` `reach_order`) orders the channels the Executive reaches a person on, for briefs (`runner.delivery_order`), messages (`message_person` and the follow-up channel pick) and nudges (`prefer_channel_for()`, which also honours availability windows): their `preferred_channel`; with `any`, the chat app they last talked in (`last_channel`, written by `note_talked` at most hourly on a Slack, Discord or private Telegram turn from their own stored id); then Slack → Discord → Telegram; then email as the backup (first for an `email` preference). Each caller still skips a channel that isn't set up or an id it can't use. Departments can also set `slack_channel_id` / `discord_channel_id` / `telegram_chat_id` to broadcast to a team room instead of DMing the head.\n\nEach person's page also lists their **open loops** (see Attunement) — what they committed to or were asked for in conversation — with a *Mark done* action; archiving a person closes their open loops, since an archived person can no longer be chased. The list is the principal's or that person's only; `people.store.is_principal_or_self` is the one check behind it, the close action and reply feedback.\n\nThe Executive can change the roster from a chat turn (`upsert_person`, `archive_person`, `set_department_head`) only when the principal asks, from the web chat, their own Slack or Discord, or a private Telegram chat when `TELEGRAM_WEBHOOK_SECRET` is set to a value Telegram can send. A teammate, an email (even one from the principal's address), Google Chat, the CLI, the MCP server and unattended runs are refused, because a team member's People row is also the web sign-in allow-list, the outbound-email allow-list and approval routing. The People page follows the same rule: adding, editing and archiving people (`POST`/`PATCH /people`, `POST /people/{id}/archive`) are the principal's alone, or anyone's while no one is principal yet, so a first setup can add its owner (`chat._caller_is_principal_or_unclaimed`). Anyone else gets 403. That includes a teammate editing their own entry: its email and chat ids are how they sign in and how the Executive reaches them, and the rest decides their approvals and how they are chased. Their working style and open loops stay theirs. A signed-in owner whose email is on no People entry is refused too; they sign in with the email on their entry, or re-run the setup interview, which fills a missing one with the email they signed in with. The one roster change that can come from an email is the principal's tokened reply to a roster-request email, and the one a chat turn can make beyond these tools is `resolve_roster_request` — both below, both the principal's alone.\n\n### Each person's notifications\n\nEvery team member sets their own on Settings → Notifications (`GET`/`PUT /notifications/settings`, see API), stored one row per person in `person_notification_settings` (`people.notifications`): **quiet hours** (off by default; from–to in their zone, which may cross midnight; every day, or weekdays with the weekend quiet all day), **delivery** — *As they happen* (the default) or *In my briefs* — and **Let urgent things through** (on by default). No row means the defaults, which hold nothing back.\n\nThey apply only to the Executive's **own** outreach: a DM sent while a proactive source is tagged (`attunement.outcomes.tag_proactive` — the scheduler's nudge dispatch, the reflection, research findings, alert review). Replies in a conversation, a message someone asked for in chat, a follow-up someone scheduled (it was asked for at a time), the briefs themselves and reminders a person set are never held. Quiet hours are deliberately not part of the outbound guard's availability check, which runs on every DM — the brief included — and would refuse a brief scheduled inside the person's own quiet hours.\n\nTwo places apply them. **Nudge routing** asks `people.channel` with `quiet_hours=True` (`is_within_availability`, `next_window_for`, `prefer_channel_for`, `next_available_window`), so the nudge engine and the scheduler's `proactive_nudge` re-check defer a nudge to the end of quiet hours exactly as they defer to an availability window — nothing is written at 3am. **Every other proactive DM** meets `outbound_guard.check_hold` in the send handlers once the anti-spam guard has allowed it: if the recipient is in quiet hours, or chose *In my briefs*, the text is stored in `held_messages` (`people.held_messages`) and the tool returns `{\"status\": \"held\"}`; `message_person` stops there instead of trying another channel; alert review counts a route as made but not a nudge toward its cap, and holds no second nudge about an alert while one waits. If it can't be stored it is sent. A scheduler dispatch whose message was held is audited as `held`, not `delivered`. Quiet-hours rows go out when quiet hours end, as one *While you were away* message (see Scheduler); brief rows ride the principal's next morning brief or end-of-day digest as one *While you were away* block, and anyone else's go out on their own when those briefs run, since only the principal has a brief. Switching back to *As they happen* releases what waited. Whoever sends a row claims it first, so a brief and a release pass never both send it. Someone on leave or outside their availability windows is postponed to when that ends; any other failed delivery backs off (2, 4, 8 … minutes, at most an hour). Rows are capped at 50 per person and dropped after 7 days undelivered, each drop audited privately to the person as a count; a brief row with no brief to carry it for 36 hours goes on its own.\n\n**Urgent** messages — approval nudges (`nudge_stalled`) and alert-review DMs about a critical (`urgent`-severity) alert (`notifications.urgent_send`) — get through when the person lets urgent things through. A workflow approval question (`gate_delivery`, `notifications.approval_send`) is never held, whatever the setting, since only a reply to it resolves the gate. If quiet hours cover every availability window, a nudge falls back to the next window rather than being dropped. Not covered: nudges on the email channel (the MCP mail tool bypasses the send handlers, so routing defers them for quiet hours but they are never held for a brief), department channel broadcasts, and the alert dispatcher's own channels (the web stream, a self-addressed email, and a Slack DM that is still a stub). Department check-ins send no DM: what they propose becomes approval cards.\n\n### Team and contacts\nA `Person` is either on the **team** (`kind = \"team\"`, the default) or a **contact** (`kind = \"contact\"`): a client, contractor or advisor the Executive may need to email or invite but who must not get what a team row grants. Before contacts existed, emailing a client meant adding them as a Person, which also let them sign in and talk to the bot.\n\nContacts are **denied by default**. `people.store.list_people`, every `find_person_by_*` (and `find_person_by_channel_ref`), and the `people.registry` cache return team members only unless the caller passes `include_contacts=True`; `find_approvers` has no opt-in at all. So a contact is never in the web sign-in list (`GET /auth/allowed-emails`), the Slack / Telegram / Discord inbound gates, approvers, the alert review's roster slice, `/today`'s people list and insights, open-loop owners (an ask of a contact opens no loop, so the nudge engine never chases them), a workflow gate's approver or cadence recipient, the recap DM after a meeting, or a department head (`set_department_head` and `PATCH /departments/{slug}` refuse one). `get_person(id)` returns either kind and its security-relevant callers check `kind`.\n\nThe roster reads that **opt in** are few: the Executive's egress checks (Gmail, Calendar and Drive in the MCP gateway, calendar attendee resolution, and the chat DM tools), the People page's listing (`GET /people?include_contacts=true`), the chat `list_people` tool (both kinds, with `kind`), the org prompt block, and the roster-request intake (which looks a sender up with contacts included only to leave a contact out). Even there, reaching a contact needs **the principal's own turn**: `people_tools.contacts_reachable_now()` is true only on a turn the principal started on a surface that verified them — the same rule as roster writes above — or inside `grant_contact_egress()`, which the decisions approve route uses when the approver resolves to the principal. An inbound email (even from the principal's address), a teammate's turn, and unattended runs (scheduler, workflows, alert review) can still reach the team but not contacts. A workflow resumed by `apply_resolution`'s kick runs with no chat session, like the resumer's poll loop, so it never inherits the approver's turn.\n\n**Contacts are private to the principal.** A solo user may be an executive whose contacts include their boss, reports, board and clients, so nobody else using the install sees one. The rule is `contacts_reachable_now()` again: off the principal's own verified turn a contact does not exist — `list_people` returns the team with the same count, a contact's id or address gets exactly the error an unknown one gets (never a contact-specific refusal), and the org prompt block has no Contacts section. The API applies the same rule by caller (`people.caller_is_principal`: the signed-in email is the principal's, or no `x-caller-email` — CLI, local login; unlike the owner rule above, nobody is while no one is principal yet): `?include_contacts=true` is ignored for anyone else, a contact's id is a 404 on `GET /people/{id}` and its open loops exactly like a missing one (editing and archiving are already a 403 before any lookup), only the principal may create a contact or turn someone into one (403, a first setup with no principal included), and `GET /people/me` tells the UI whether to offer contacts at all.\n\nWhatever the Executive raises about a contact stays with the principal too. A meeting with a contact is proposed to the principal (not the scope's approver), its decision is hidden from everyone else's `/decisions` (404 by id), and its card carries `alerts.models.PRIVATE_ALERT_TAG`. So does an alert raised on a turn about the principal's private mail (`Session.private_to_principal`, set by the email poller for mail from a contact and for mail the principal forwards): the pipeline routes it to the principal, stores it without a live push or broadcast, and dedups it only against other private alerts. Private alerts appear on the principal's own `/today` and chat digest only — never a teammate's, the activity rail, the narrative, the morning brief, the end-of-day digest, the reflection or the alert review — and a teammate's ack, reopen or bulk sweep cannot touch one (404). A private turn itself reaches the principal and nobody else (email, DMs, invites). It is not offered, and cannot run, the tools that post to other people, publish where the team reads (artifacts, goal updates, saved workflows, skill drafts) or start work outside the turn (`schedule_tools.PRIVATE_TURN_WITHHELD_TOOLS`), and of the MCP tools it may use only Google Workspace reads and the Gmail send the gateway narrows to the principal. Its audit rows are the principal's alone, as is any row on the principal's own turn that names one of their contacts (see Audit Log). DMs to a contact stay off the activity rail.\n\n`upsert_person` takes `kind`: explicit wins; an update that omits it keeps the row's kind (the store's full-row UPDATE preserves it, so nothing promotes a contact to the team by accident); a new person defaults to `team`, or to `contact` in solo mode. On the principal's own verified turn the org prompt block lists contacts in a short `## Contacts` section after the team block (or, in solo, after `## Your Principal` and their goals) — name, role/company and whether an email is on file, capped at 25 — in both modes. Every other turn gets the team block, byte-identical to an install with no contacts, so the 5-minute cached block has two stable variants.\n\nOn the **People page** the principal sees **Team** and **Contacts** tabs; anyone else sees the team only (no Contacts tab, no kind selector). Solo mode opens on Contacts, and its Team tab shows only the principal. The add form has a kind selector and hides authority scopes, reply SLA and the principal checkbox for a contact; a contact's page hides scopes, SLA, availability windows, open loops, outreach and working style. Adding a team member in solo offers to switch the workspace to team mode, or keep solo.\n\n### Contacts the Executive keeps up to date\n`people.contact_keeper` keeps the principal's contacts current on its own, from **their own mailbox** under Act as me (`delegation.gmail_for`: Gmail or Outlook, never the MCP gateway). It runs in the background (the scheduler starts it every 6 hours), never in a chat turn, and only while the principal has Act as me on with their own mailbox connected. The switch **Keep my contacts up to date** (`contact_keeper_state.mode`: `on`, the default; `ask`; `off`) sits on the Contacts tab.\n\n- **Adds**: someone the principal emailed at least twice in 60 days (sent mail they wrote: no lists, bulk or auto mail, nothing Handle it for me sent, at most 10 recipients) or met (meetings from 30 days back to 14 ahead that they organised or accepted, at most 10 guests). Names come only from To/Cc headers and guest lists, never a mail's text. Skipped: their own and the Executive's addresses, anyone in People (archived included), role and no-reply addresses, company-domain addresses (`identity.is_company_address`), and anything blocked. At most 10 per run.\n- **Updates**: an alias when a known contact (two-word full name) is mailed twice at a new address under that exact name as the principal's To line wrote it, always as a suggestion whatever the switch (mail to the contact may go there, so one reply to a lookalike must not attach it; an undone or archived address never is); a fuller name for a one-word contact; a role (\"Head of Procurement, Acme\") read from the contact's newest mail (`latest_from`) when its sender is authenticated and the From address is theirs. The role read is one forced-tool model call on the mail's last 1,200 characters, kept only if every word is in the mail, it is at most 10 words and 80 characters, and it carries no link, address or instruction. At most 5 role reads per run and 20 a day, and an address again only after 14 days. A field the principal typed is never overwritten: a name or role is written only when empty or last set by the keeper.\n- **Record and Undo** (`people.contact_changes`): every change is a row (`contact_changes`) with its reason and evidence. Under `ask` it is stored as a suggestion and applied on **Add**. **Undo** archives an added contact or restores the old value if nobody changed it since; Undo and **Not now** block that address, field and value (`contact_keeper_blocks`) so they are never proposed again. A contact added from chat is recorded too (reason `asked`), so it has an Undo. Each run writes one private `contacts_kept` audit row with counts, no names.\n- **Contact details**: the same signature read copies the sender's other contact details (`Person.details`, table `person_details`: phone, mobile, website, LinkedIn, another profile, postal address). Each is kept only when the mail writes it (a phone's digits as one number, every word of an address, a link's host and path), a website only on the sender's own mail domain (never a free mail one) and a profile only on a known site (`people.contact_info`). One of each kind is the keeper's: a newer one replaces it, and a kind anyone typed a value for is left alone (no second phone next to a typed one). Each is a change with **Undo**. People and chat (`upsert_person`'s `details`) can type them too. The people routes and `list_people` return a person's details only to the principal and to that person; everyone else gets the row with `details` empty. They are for the reader only: nothing is ever sent to a phone or link.\n- **Team members** get the same signature read (role when empty, contact details) and the same note; their name and addresses are never changed.\n- **What it remembers** (`people.contact_notes`, table `contact_notes`): for every contact and team member, when the principal last emailed, heard from (the date of their newest authenticated mail) and met them, their next meeting, counts, and a short note with up to 6 topics on what they deal with each other about. The note is written by one forced-tool call only from the principal's **own** sent mail to that contact (their new text, without quoted mail; nothing Handle it for me sent) and the titles of meetings they organised, so nobody else's words go in. It is rewritten only when that mail or those meetings change, at most 5 a pass and 20 a day, and a note holding a link, address or tag is dropped. \"Last heard from\" is read for up to 20 people a pass, whatever the note budget. The Executive reads it on the principal's own turn in `list_people` (contacts and teammates) and `ask_about_person` (contacts), to suggest who to ask; the person's page shows it to the principal as **What I remember**. A teammate's note is the principal's alone too: it comes from the principal's mailbox, never the teammate's own mail to the Executive or their peer memory. Archiving someone deletes it.\n\nAll of it is the principal's alone: the routes answer only `caller_is_principal`, and the rows are private. An added contact is an ordinary contact, with the rules above.\n\n### Addresses: aliases and company domains\nA `Person` has one **primary** address (`email`) — the one they sign in with — and up to 10 **aliases** (`email_aliases`, the `person_emails` table: `email_norm` unique across the roster, `source` `manual` / `chat_tool` / `roster_request`, deleted with the person or when they are archived, so an archived person never blocks the address). An alias matches their mail and may be emailed, but **is never a login**: `find_person_by_email` stays primary-only and is what web sign-in and `/auth/allowed-emails` read; mail goes through `find_person_by_address` (primary or alias, case-insensitive, the team row winning over a contact) and `people.identity.resolve_email_sender`. Aliases are edited on the person's page (\"Other addresses\") and by `upsert_person`'s `email_aliases` (the full list, replacing the current one); an address that is already someone else's primary or alias is refused (`AddressInUseError`, a 409 on the API), and so is a new primary that is someone's alias. Onboarding still imports no addresses beyond the owner's own sign-in email, and an alias counts as taken there too.\n\nOn top of exact matches, `resolve_email_sender` applies a **company-domain rule, team only**: on one of the company's own domains an address's local part (lowercased, `+tag` dropped) matches a teammate whose own address on any company domain has that local part — `[email protected]` is the Anna whose address is `[email protected]`. A domain alone proves nothing: an unknown local part on the company domain matches nobody (it becomes a roster request, pre-filled as a teammate), and when two teammates would match, nobody does. The company's domains are the `company_domains` workspace setting (Settings → Workspace, `PUT /workspace`, at most 10), or, when that is unset, the domains of the principal's own addresses — never a free-mail provider's (`identity.FREEMAIL_DOMAINS`: gmail.com, outlook.com and the like, refused as a setting too), where a local part says nothing about who someone is. The outbound gates build `RosterAllow` from the same rule, so the Executive may email exactly the addresses inbound mail would match (see Integrations).\n\n### New senders: roster requests\nWhen someone not on the roster — and not one of the principal's contacts — writes in (an email address, a Slack or Discord user, a private verified Telegram chat), their message is held as a **roster request** (`people.roster_requests`: tables `roster_requests`, `roster_request_messages`, `roster_ack_log`) instead of being dropped; they get a fixed acknowledgement (an email sender only when Gmail authenticated them — see Integrations) and the principal gets a private card on `/today`, which says whether they were actually told (`surface_card(acknowledged=...)`, from `ack_sent_at` after the acknowledgement is attempted) and, for a chat or company-domain sender, a prompt on their delivery channel (the per-channel mechanics are in Integrations). A request holds at most 10 messages (later ones only count), is one per `(channel, sender)` while pending, and suggests a match only from a chat account whose profile email is already someone's (never from a display name, which the sender chooses) — and `suggested_kind = team` for an address (or a chat profile email) on a company domain.\n\nThe principal answers one of three ways, all ending in `roster_requests.resolve` (a claim on the row, so two answers race and one wins; nothing is written unless it succeeds):\n\n- **approve** — add them as a new Person with this address or chat id, `team` or `contact`;\n- **link** — they are someone already on the list writing from somewhere new: an email becomes that person's alias — never their primary, even if they had none, so a link never creates a login, a chat id fills that person's field (replacing a different one only when the principal says so; otherwise `ChannelIdConflict`);\n- **decline** — leave them off; they are not raised again for 30 days.\n\nOnly the principal answers, and only where they are verified: the **web card** (`POST /people/requests/{id}/approve|decline`, which requires an explicit team/contact choice), their **own Slack or Discord DM or verified private Telegram chat** (the `resolve_roster_request` chat tool, which answers only ids the server showed that turn in its `<roster_requests>` block — see Agents), or an **email reply from their own primary address** (not an alias) carrying the request's one-time token, which the gateway hides from every other read of the Executive's mailbox so no model turn can copy it. When the chat or email answer does not say team or contact, the default is team for a sender on a company domain, else contact — the least access. Whoever answers, the held messages are then replayed as if the sender had been on the roster all along (a new contact gets only held email, on the private contact path); a decline drops them. A roster write outside a request — the People page (`POST`/`PATCH /people`) or `upsert_person` — reconciles too: any pending request whose sender now matches someone closes as `superseded` and replays.\n\nUnanswered requests expire after `ROSTER_REQUEST_TTL_DAYS` (14). A held message's text is kept only until the request is resolved or expires (then purged, even on decline), its card, routes and audit rows are private to the principal (404 for anyone else), and the tables are wiped with the rest of the People data on a company swap or fixture load.\n","mermaid": "flowchart TD\n People[\"People registry (Person, is_principal, kind)\"]\n Contacts[\"Contacts (kind = contact): egress only on the principal's turn\"]\n Req[\"Roster requests (new senders held for the principal)\"]\n Dept[\"Departments (8 seeded once)\"]\n GC[\"Goals / checklists / cadences\"]\n Head[\"Head + voice persona\"]\n Gate[\"Authority gate (propose_only vs autonomous)\"]\n Scope[\"AuthorityScope to find_approvers\"]\n Chan[\"Channel resolution (Discord / Telegram / email)\"]\n Dept --> GC\n Dept --> Head\n Dept -->|proposes action| Gate\n Gate -->|required_scope| Scope\n Scope -->|ranks approvers| People\n People --> Chan\n People -. include_contacts .-> Contacts\n Req -->|approve / link| People\n Dept -->|team room| Chan\n class People storage\n class Contacts storage\n class Req storage\n class Dept storage\n class GC storage\n class Gate compute\n class Scope compute\n class Chan compute\n class Head compute",- "generated_at": "2026-10-21T00:00:00Z"+ "generated_at": "2026-10-21T16:00:00Z"}diff --git a/packages/core/openexecutive/guide/prebuilt/company_profile.json b/packages/core/openexecutive/guide/prebuilt/company_profile.jsonindex 99b396ee..443609f6 100644--- a/packages/core/openexecutive/guide/prebuilt/company_profile.json+++ b/packages/core/openexecutive/guide/prebuilt/company_profile.json@@ -1,7 +1,7 @@{"section_id": "company_profile","title": "Company Profile & Onboarding",- "markdown": "**What it is:** Your company's identity and strategy (mission, market, competitors, financials, and org structure) that grounds everything the Executive says. When you use Open Executive just for yourself, it is your **Business profile** (if you own the business) or **Your work**, with the same sections.\n\n**What it does:**\n\n- **Sets up the Executive from your own words** the first time: describe your work, optionally attach a deck, one-pager or CV, and it works out the rest. It asks only about what's missing, then shows you a filled-in draft profile, team list, and departments.\n- Lets you review and edit that draft before anything is saved. Nothing is written until you click **Save & finish setup**. A step-by-step form is linked at the bottom of the setup screen.\n- Lets you view and edit every section of the profile any time afterward.\n- Feeds this context to all specialists so advice fits *your* company, not a generic one.\n- After setup, runs background research and suggests things worth watching.\n\n**How to use:**\n\n*First-time setup*\n\n1. On a new install the app opens setup for you. You can also click **Set up profile →** on the banner on Home, or **Set up company** in the sidebar before a profile exists. Setup opens at `/onboard`, full screen.\n2. **Set up your Executive**: describe your work in the box (what you do, who you sell to, roughly how big you are, what you're focused on this year). Optionally click **+ Attach a one-pager, deck or CV (optional)** (up to 8 files: PDF, Word, Excel, CSV, Markdown or text). Click **Continue**. **Skip, I'll answer questions instead** goes straight to the questions. **I'll do this later** takes you to Home, which keeps setup one tap away.\n3. **Here's what I got** (or **Who is this for?** if you skipped the description): pick **Personal** (your own executive) or **Team** (an executive for the whole company, with a lead for each department) and check what it read: your company and this year's focus, and for Personal your role, title and who you report to. Fix anything that's off and click **Continue**. This saves the mode and your role, which you can change later in **Settings → Workspace**.\n4. Answer the follow-up questions (up to 5; the screen shows how many have been asked). Press Enter to send. Click **Skip ahead — draft my profile now** to stop early.\n5. **Here's what I understood**: edit each profile section. For a team, edit **Your team** (**Add person**, remove someone, tick **This is me** on exactly one person) and **Departments** (**Add department**, each with what it owns, a head and an authority level). For Personal, enter your own name and title. Check **Your sign-in email**: it's filled in with the account you're signed in with, and it's how the Executive knows the person marked as you is you. If you're setting this up for someone else, enter the email they will sign in with. Anything it couldn't work out is listed under **I couldn't determine these**.\n6. Click **Save & finish setup**, or **Not quite — ask me more questions** to go back to the conversation.\n7. **How should your Executive sound?**: pick a voice, or **Skip for now**. You can change it later in Settings.\n8. **Your Executive is ready**: a first question, built from your priorities, is filled in under **Start here**. Edit it and click **Ask your first question**, or **Open Home instead**. The same screen links to connecting the Executive's own email account and letting it read your own mailbox.\n\nA setup in progress survives a refresh or a closed tab: open setup again and it picks up at the same question or draft. Click **Start over** on the \"Picking up where you left off\" note to begin again from the description.\n\nTo change an earlier answer, click **Edit** next to it under **Your answers** (or under **Built from your answers** on the draft), change it and click **Save and continue**. The questions after it are asked again, or the profile is redrafted if it already has enough. Files you attached to the description are kept.\n\nPrefer filling in fields? Click **Prefer a form? Use the step-by-step version** for a wizard of up to 12 questions (fewer for Personal). Later steps are optional and can be skipped.\n\n*Afterwards*\n\n- Open **Company** in the sidebar and pick the **Company profile** tab. Click **Edit** on a section, make your changes, and click **Save** (or **Cancel**). List fields take one item per line. The sections are **Company Basics**, **Mission & Vision**, **Target Customer**, **Competitive Landscape**, **External Dependencies**, **Strategic Priorities**, **Culture & Values**, **Org Structure** and **Financials**.\n- **External Dependencies** holds vendors and tickers. A vendor or ticker named there counts as company data: the Executive starts watching its status page or filings on its own, without asking you first.\n- **Re-run setup**, in the **⋯** at the top of the profile, takes you back through the conversation. A re-run skips the Personal / Team step. Only the current owner can mark someone else as **This is me** when re-running it, because saving makes that person the owner.\n- The help panel (the sparkle button at the top right) can fill in profile sections for you. The section opens with the suggested values and you still click **Save**.\n\n**Who can change it:** anyone signed in can edit the profile sections. Saving setup adds people and makes someone the owner, so once there is an owner only they can hand that role to someone else. The step-by-step form can only be saved by the owner once there is one.\n\n**Example description:** *\"We're Northwind Tools — a 40-person B2B SaaS company selling inventory software to independent hardware stores in the US. About $6M ARR, growing 60% a year, Series A last spring. Our main competitors are Stockly and spreadsheets. This year's priorities are launching in Canada and getting net revenue retention above 110%.\"*\n\n*Note: until a profile exists, Home shows a banner that responses will be generic, with **Set up profile →**.*",+ "markdown": "**What it is:** Your company's identity and strategy (mission, market, competitors, financials, and org structure) that grounds everything the Executive says. When you use Open Executive just for yourself, it is your **Business profile** (if you own the business) or **Your work**, with the same sections.\n\n**What it does:**\n\n- **Sets up the Executive from your own words** the first time: describe your work, optionally attach a deck, one-pager or CV, and it works out the rest. It asks a few quick questions about what's missing, then shows you a filled-in draft: the profile, your team, this year's goals with an owner on each, and departments only if your description already listed people by department.\n- Lets you review and edit that draft before anything is saved. Nothing is written until you click **Save and start**. A step-by-step form is linked at the bottom of the setup screen.\n- Lets you view and edit every section of the profile any time afterward.\n- Feeds this context to all specialists so advice fits *your* company, not a generic one.\n- After setup, researches your market and competitors and puts what it finds in your brief, and checks your goals every week.\n\n**How to use:**\n\n*First-time setup*\n\n1. On a new install the app opens setup for you. You can also click **Set up profile →** on the banner on Home, or **Set up company** in the sidebar before a profile exists. Setup opens at `/onboard`, full screen.\n2. **Set up your Executive**: describe your work in the box (what you do, who you sell to, roughly how big you are, what you're focused on this year). Optionally click **+ Attach a one-pager, deck or CV (optional)** (up to 8 files: PDF, Word, Excel, CSV, Markdown or text). Click **Continue**. **Skip, I'll answer questions instead** goes straight to the questions. **I'll do this later** takes you to Home, which keeps setup one tap away.\n3. Setup works out whether this is **Personal** (just you) or for your **Team** from what you wrote, without asking: Team when you named colleagues at your company, Personal otherwise. For Personal it also keeps the role, title and who you report to that you mentioned.\n4. **A few quick questions**: up to four, each with suggested answers. Tap one, type your own and press Enter, or click **Skip**. Every question is optional. Click **Draft it now** to stop early.\n5. **Here's** *your company*: the first line says who it's set up for, for example **Set up for your team: Dana, Sam, Priya, Lee**. Click **Change to just me** or **Change to a team** to switch. Edit the profile sections, and **This year**: your goals, each with an owner (**Add a goal**, pick an owner, or ✕ to remove one). For a team, edit **Your team** (**Add person**, remove someone, tick **This is me** on exactly one person). **Departments** appear only when your description already listed people by department. For Personal, enter your own name and title. Check **Your sign-in email**: it's filled in with the account you're signed in with, and it's how the Executive knows the person marked as you is you. If you're setting this up for someone else, enter the email they will sign in with. **Please check** lists anything it guessed. **After you save, I'll** says what happens next.\n6. Click **Save and start**, or **Back to questions** to go back to the conversation. Saving takes you to Home, where the getting-started checklist shows what's worth connecting next. Pick how the Executive sounds any time in Settings.\n\nA setup in progress survives a refresh or a closed tab: open setup again and it picks up at the same question or draft. Click **Start over** on the \"Picking up where you left off\" note to begin again from the description.\n\nTo change an earlier answer, click **Edit** next to it under **Your answers** (or under **Built from your answers** on the draft), change it and click **Save and continue**. The questions after it are asked again, or the profile is redrafted if it already has enough. Files you attached to the description are kept.\n\nPrefer filling in fields? Click **Prefer a form? Use the step-by-step version** for a wizard of up to 12 questions (fewer for Personal). Later steps are optional and can be skipped.\n\n*Afterwards*\n\n- Open **Company** in the sidebar and pick the **Company profile** tab. Click **Edit** on a section, make your changes, and click **Save** (or **Cancel**). List fields take one item per line. The sections are **Company Basics**, **Mission & Vision**, **Target Customer**, **Competitive Landscape**, **External Dependencies**, **Strategic Priorities**, **Culture & Values**, **Org Structure** and **Financials**.\n- **External Dependencies** holds vendors and tickers. A vendor or ticker named there counts as company data: the Executive starts watching its status page or filings on its own, without asking you first.\n- **Re-run setup**, in the **⋯** at the top of the profile, takes you back through the conversation. A re-run keeps your Personal or Team setting. Only the current owner can mark someone else as **This is me** when re-running it, because saving makes that person the owner.\n- The help panel (the sparkle button at the top right) can fill in profile sections for you. The section opens with the suggested values and you still click **Save**.\n\n**Who can change it:** anyone signed in can edit the profile sections. Saving setup adds people and makes someone the owner, so once there is an owner only they can hand that role to someone else. The step-by-step form can only be saved by the owner once there is one.\n\n**Example description:** *\"We're Northwind Tools — a 40-person B2B SaaS company selling inventory software to independent hardware stores in the US. About $6M ARR, growing 60% a year, Series A last spring. Our main competitors are Stockly and spreadsheets. This year's priorities are launching in Canada and getting net revenue retention above 110%.\"*\n\n*Note: until a profile exists, Home shows a banner that responses will be generic, with **Set up profile →**.*","mermaid": null,- "generated_at": "2026-10-17T15:48:15Z"+ "generated_at": "2026-10-21T16:00:00Z"}diff --git a/packages/core/openexecutive/onboarding/commit.py b/packages/core/openexecutive/onboarding/commit.pyindex e18bef3e..7752f338 100644--- a/packages/core/openexecutive/onboarding/commit.py+++ b/packages/core/openexecutive/onboarding/commit.py@@ -14,7 +14,9 @@ match drafted departments onto existing ones and create only what isgenuinely new.A new install has no departments until this step runs, so the draft decides-which ones exist: none for a solo workspace or a team that names none. A new+which ones exist: none for a solo workspace, and none unless the user's own+words listed people by department. Goals stand on their own: the draft's goals+are saved with an owner, inside a department only when one was drafted. A newdepartment named after a specialist's area ("Finance", "HR") is wired to thatspecialist, the same rule the Executive's ``create_goal`` tool uses, so itsgoals get check-ins.@@ -24,13 +26,14 @@ from __future__ import annotationsimport loggingimport reimport sqlite3+from datetime import datetimefrom typing import TYPE_CHECKING, Anyfrom openexecutive.utils.slug import DEPARTMENT_SLUG_FALLBACK, slugifyif TYPE_CHECKING: # pragma: no cover - typing onlyfrom openexecutive.memory.company_profile import CompanyProfile- from openexecutive.onboarding.interview import DepartmentDraft, PersonDraft+ from openexecutive.onboarding.interview import DepartmentDraft, GoalDraft, PersonDraftfrom openexecutive.people.models import AuthorityScopelogger = logging.getLogger(__name__)@@ -489,16 +492,92 @@ def reconcile_onboarding_departments(logger.warning("onboarding: department registry invalidate failed")+def save_onboarding_goals(+ drafts: list[GoalDraft],+ person_ids: dict[str, int],+ principal_id: int | None,+) -> int:+ """Add the drafted goals for this year. Returns how many were created.++ Additive like the departments: a goal whose wording matches one already+ set for this year or ongoing (any department, case-insensitive) is left+ alone, so re-running setup does not duplicate goals. A goal's owner is the named+ person, else the principal; its department is the drafted one it names,+ matched the way reconcile_onboarding_departments matched it, else none.+ Best-effort: logs the exception type only and keeps what was saved.+ """+ created = 0+ try:+ from openexecutive.departments.store import (+ initialize_db as init_departments_db,+ )+ from openexecutive.departments.store import (+ insert_goal,+ list_all_goals,+ list_departments,+ match_department,+ )+ from openexecutive.memory.workspace_settings import get_user_timezone++ named = [g for g in drafts if g.title.strip()]+ if not named:+ return 0+ init_departments_db()+ known = [d.config for d in list_departments()]+ year = str(datetime.now(get_user_timezone()).year)+ # Only this year's goals (any period label naming the year, e.g.+ # "2026", "Q3 2026") and ongoing ones count: last year's goal with+ # the same wording is a new goal this year.+ seen = {+ g.key_result.strip().lower()+ for g in list_all_goals()+ if g.period_type == "ongoing" or year in g.period_value+ }+ ids = {name.strip().lower(): pid for name, pid in person_ids.items()}++ for draft in named:+ title = draft.title.strip()+ if title.lower() in seen:+ continue+ seen.add(title.lower())+ owner = ids.get(draft.owner_name.strip().lower()) if draft.owner_name.strip() else None+ department = (+ match_department(draft.department_title.strip(), known)+ if draft.department_title.strip()+ else None+ )+ insert_goal(+ department.slug if department is not None else None,+ period_type="year",+ period_value=year,+ key_result=title,+ target=draft.target.strip(),+ # Explicit even for the principal: a goal in a department+ # would otherwise fall to its head, not to whoever was named.+ owner_person_id=owner if owner is not None else principal_id,+ )+ created += 1+ return created+ except Exception as exc:+ logger.warning(+ "save_onboarding_goals failed (%s) after %d goal(s)", type(exc).__name__, created+ )+ return created++def derive_org_structure(profile: CompanyProfile,people: list[PersonDraft],departments: list[DepartmentDraft],+ goals: list[GoalDraft] | None = None,) -> CompanyProfile:- """Fill ``profile.org_structure`` from the people and department drafts.+ """Fill ``profile.org_structure`` from the people and department drafts,+ and this year's priorities from the goals when there are any.- The onboarding review screen hides the Org Structure section precisely so- this stays a derived value — otherwise the user would have two places to- edit the same list and they would silently diverge.+ The onboarding review screen hides the Org Structure and priorities+ sections precisely so these stay derived values — otherwise the user+ would have two places to edit the same list and they would silently+ diverge."""titles = [d.title.strip() for d in departments if d.title.strip()]leadership = [@@ -509,4 +588,10 @@ def derive_org_structure(org = profile.org_structure.model_copy(update={"departments": titles, "leadership_team": leadership})- return profile.model_copy(update={"org_structure": org})+ update: dict[str, Any] = {"org_structure": org}+ goal_titles = [g.title.strip() for g in goals or [] if g.title.strip()]+ if goal_titles:+ update["strategic_priorities"] = profile.strategic_priorities.model_copy(+ update={"current_year": goal_titles}+ )+ return profile.model_copy(update=update)diff --git a/packages/core/openexecutive/onboarding/interview.py b/packages/core/openexecutive/onboarding/interview.pyindex 7d2eb7e6..577369a5 100644--- a/packages/core/openexecutive/onboarding/interview.py+++ b/packages/core/openexecutive/onboarding/interview.py@@ -1,8 +1,9 @@"""Conversational company setup — the interview loop behind ``/onboard``.The user describes their business in their own words; this module asks-clarifying questions until it can draft a company profile, a leadership-roster, and a department list. Nothing is written anywhere: ``advance``+a few quick questions (each with suggested answers to tap) until it can draft+a company profile, the people, this year's goals with an owner on each, and+departments when the user's own words already list people by department. Nothing is written anywhere: ``advance``returns either a ``Question`` or a ``CompanyDraft``, and the route owns boththe session state and the single write (see ``onboarding/commit.py``).@@ -53,8 +54,8 @@ from openexecutive.utils.slug import DEPARTMENT_SLUG_FALLBACK, slugifylogger = logging.getLogger(__name__)# Budgets. The user can always short-circuit with force_draft, so these only-# bound a runaway model.-MAX_QUESTIONS = 5+# bound a runaway model. Setup asks two to four quick questions.+MAX_QUESTIONS = 4# Must stay above what a single legal /start can produce — a 20k description# plus 8 attachments at 15k extracted chars each is ~140k. A lower ceiling# locked the user out of the conversation on turn one for doing exactly what@@ -80,6 +81,13 @@ _REPAIR_ECHO_CHARS = 2000_MAX_PROSE_QUESTION_CHARS = 1000+# Suggested answers shown under a question as buttons to tap.+MAX_OPTIONS = 4+MAX_OPTION_CHARS = 120+# This year's goals in a draft. More than this is a plan, not setup.+MAX_GOALS = 12+MAX_GOAL_CHARS = 300+ASK_TOOL_NAME = "ask_clarifying_question"EMIT_TOOL_NAME = "emit_company_draft"@@ -143,10 +151,30 @@ class DepartmentDraft(BaseModel):authority_level: AuthorityLevel = AuthorityLevel.PROPOSE_ONLY+class GoalDraft(BaseModel):+ """One of this year's goals. Saved with an owner and, only when the+ draft has departments, optionally inside one (see commit.py)."""++ model_config = ConfigDict(extra="ignore")++ title: str+ target: str = ""+ # A person's full_name from the draft; empty = the principal.+ owner_name: str = ""+ # A drafted department's title; empty = no department.+ department_title: str = ""++ @field_validator("target", "owner_name", "department_title", mode="before")+ @classmethod+ def _no_null(cls, value: Any) -> Any:+ return "" if value is None else value++class CompanyDraft(BaseModel):profile: CompanyProfilepeople: list[PersonDraft] = Field(default_factory=list)departments: list[DepartmentDraft] = Field(default_factory=list)+ goals: list[GoalDraft] = Field(default_factory=list)confidence_notes: list[str] = Field(default_factory=list)summary: str = ""@@ -154,6 +182,8 @@ class CompanyDraft(BaseModel):class Question(BaseModel):question: strhint: str = ""+ # Suggested answers the user can tap instead of typing.+ options: list[str] = Field(default_factory=list)@field_validator("hint", mode="before")@classmethod@@ -162,6 +192,20 @@ class Question(BaseModel):# reason to throw the whole question away.return "" if value is None else value+ @field_validator("options", mode="before")+ @classmethod+ def _tidy_options(cls, value: Any) -> Any:+ # Lenient like the hint: a bad option list costs the buttons, never+ # the question. Blank, overlong and repeated options are dropped.+ if not isinstance(value, list):+ return []+ out: list[str] = []+ for item in value:+ text = item.strip() if isinstance(item, str) else ""+ if text and len(text) <= MAX_OPTION_CHARS and text not in out:+ out.append(text)+ return out[:MAX_OPTIONS]+@dataclass(frozen=True)class DraftError:@@ -180,10 +224,10 @@ class DraftError:_ASK_TOOL: dict[str, Any] = {"name": ASK_TOOL_NAME,"description": (- "Ask the user ONE clarifying question about their company. Use this "- "only while something material is still missing — the user reviews and "- "edits the draft before it is saved, so an early draft beats a long "- "interview."+ "Ask the user ONE quick question about their company, with suggested "+ "answers they can tap. Use this only while something material is "+ "still missing — the user reviews and edits the draft before it is "+ "saved, so an early draft beats a long interview."),"input_schema": {"type": "object",@@ -200,6 +244,16 @@ _ASK_TOOL: dict[str, Any] = {"you're asking, or an example answer."),},+ "options": {+ "type": "array",+ "items": {"type": "string"},+ "maxItems": MAX_OPTIONS,+ "description": (+ "Two to four short suggested answers, shown as buttons to "+ "tap. Base them on what the user already said. The user "+ "can always type their own answer or skip."+ ),+ },},},}@@ -207,12 +261,12 @@ _ASK_TOOL: dict[str, Any] = {_EMIT_TOOL: dict[str, Any] = {"name": EMIT_TOOL_NAME,"description": (- "Emit the draft company profile, leadership roster, and department "- "list for the user to review and edit. Exactly one person must have "- "is_principal=true (the user). Every department's head_person_name "- "must exactly match a person's full_name or be empty. Leave anything "- "the user did not state null or empty and name it in confidence_notes "- "— never guess a number."+ "Emit the draft company profile, people, this year's goals and (only "+ "when the user listed people by department) departments for the user "+ "to review and edit. Exactly one person must have is_principal=true "+ "(the user). Every head_person_name and owner_name must exactly match "+ "a person's full_name or be empty. Leave anything the user did not "+ "state null or empty — never guess a number."),"input_schema": {"type": "object",@@ -330,7 +384,8 @@ _EMIT_TOOL: dict[str, Any] = {"departments": {"type": "array","description": (- "The functions this business actually runs. Titles must be distinct. "+ "Only when the user's own words or files already list people "+ "by department; otherwise empty. Titles must be distinct. ""head_person_name must match a person's full_name exactly, or be empty."),"items": {@@ -348,10 +403,49 @@ _EMIT_TOOL: dict[str, Any] = {},},},+ "goals": {+ "type": "array",+ "maxItems": MAX_GOALS,+ "description": (+ "What the user wants to achieve this year, one goal per item, "+ "in their words. Also list them in "+ "profile.strategic_priorities.current_year."+ ),+ "items": {+ "type": "object",+ "required": ["title"],+ "properties": {+ "title": {"type": "string", "description": "e.g. 'Launch the hosted version'."},+ "target": {+ "type": "string",+ "description": "How they will know it is done, if they said. Else empty.",+ },+ "owner_name": {+ "type": "string",+ "description": (+ "The person's full_name who owns it. Empty means the "+ "user. Name someone else only when the user did."+ ),+ },+ "department_title": {+ "type": "string",+ "description": (+ "A drafted department's title when the goal clearly "+ "belongs to it. Else empty."+ ),+ },+ },+ },+ },"confidence_notes": {"type": "array","items": {"type": "string"},- "description": "One short line per field you could not determine, saying what is missing.",+ "description": (+ "Things the user should check after reading the draft, one plain "+ "sentence each (e.g. 'I guessed your main competitors from your "+ "site.'). Never use field names. Leave out anything the user chose "+ "not to share."+ ),},"summary": {"type": "string",@@ -432,6 +526,25 @@ def validate_draft(draft: CompanyDraft) -> list[DraftError]:)names = {n.lower() for n in all_names}+ titles = {d.title.strip().lower() for d in draft.departments if d.title.strip()}+ if len(draft.goals) > MAX_GOALS:+ add(f"That's a lot of goals for setup (limit {MAX_GOALS}). Keep the main ones.")+ for g in draft.goals:+ if not g.title.strip():+ add("Every goal needs a name.", "every goal needs a non-empty title")+ if len(g.title) > MAX_GOAL_CHARS or len(g.target) > MAX_GOAL_CHARS:+ add(f"A goal is too long (limit {MAX_GOAL_CHARS} characters).")+ if g.owner_name.strip() and g.owner_name.strip().lower() not in names:+ add(+ "A goal is owned by someone who isn't on the team list.",+ f"goal '{g.title}' owner_name '{g.owner_name}' is not in the roster",+ )+ if g.department_title.strip() and g.department_title.strip().lower() not in titles:+ add(+ "A goal is in a department that isn't on the list.",+ f"goal '{g.title}' department_title '{g.department_title}' "+ "is not a drafted department",+ )for d in draft.departments:if not d.title.strip():add("Every department needs a name.", "every department needs a non-empty title")@@ -526,7 +639,8 @@ def replay_transcript(# interpolated into them._SOLO_DRAFT_RULE = ("In the draft, `people` must hold only them — exactly one person, with "- "is_principal true — and `departments` must be empty."+ "is_principal true — `departments` must be empty, and every goal's "+ "owner_name is empty (theirs).")# No role given (or "other"): find out which kind of principal this is.@@ -548,8 +662,8 @@ SOLO_ROLE_HINTS: dict[str, str] = {"they own and run the business. "+ _SOLO_DRAFT_RULE+ " Do not ask about a leadership team or departments; ask instead "- "about what they offer, who their customers are, their pricing, their "- "cash and runway, and their top goals."+ "about what they offer, who their customers are, their pricing, and "+ "their top goals."),"in_house": ("This person is setting up Open Executive just for themselves. They "diff --git a/packages/core/openexecutive/onboarding/interview_store.py b/packages/core/openexecutive/onboarding/interview_store.pyindex bf53d6dc..fe55a996 100644--- a/packages/core/openexecutive/onboarding/interview_store.py+++ b/packages/core/openexecutive/onboarding/interview_store.py@@ -71,6 +71,7 @@ def to_payload(session: InterviewSession) -> dict[str, Any]:"questions_asked": session.questions_asked,"draft": session.draft.model_dump(mode="json") if session.draft is not None else None,"last_hint": session.last_hint,+ "last_options": list(session.last_options),"saved": session.saved,}@@ -86,6 +87,7 @@ def from_payload(payload: dict[str, Any]) -> InterviewSession:questions_asked=int(payload.get("questions_asked", 0)),draft=CompanyDraft.model_validate(draft) if draft else None,last_hint=str(payload.get("last_hint", "")),+ last_options=[str(o) for o in payload.get("last_options") or [] if isinstance(o, str)],saved=bool(payload.get("saved", False)),)diff --git a/packages/core/openexecutive/onboarding/understand.py b/packages/core/openexecutive/onboarding/understand.pyindex a06e962e..39442e7d 100644--- a/packages/core/openexecutive/onboarding/understand.py+++ b/packages/core/openexecutive/onboarding/understand.py@@ -3,12 +3,13 @@Setup now starts with the free-text box. Before the interview begins, this onemodel call reads that text (and any attached files, already extracted to text)and reports what it can already tell: personal or team, the user's role, who-they report to, the company and this year's focus. The UI shows it for-confirmation and the interview then asks only about what is missing.+they report to, the company and this year's focus. Nobody is asked "who is+this for?": the UI saves the mode read here (Personal when it is unclear) and+the role through PUT /workspace, and the review screen offers a Change link.Same epistemics as the interviewer: extract, never invent. Anything the text-does not say comes back null and the UI asks for it. Nothing here writes-anything; the UI saves the confirmed values through PUT /workspace.+does not say comes back null, and the interview asks for what matters. Nothing+here writes anything."""from __future__ import annotations@@ -41,10 +42,10 @@ SYSTEM = (f"{TOOL_NAME} exactly once.\n\n""Rules:\n""- Extract, never invent. A field the text does not state is null.\n"- "- mode: 'solo' when the user is setting this up for their own work "- "(an individual, a manager, an advisor, a founder describing themselves); "- "'team' when they describe setting it up for a whole company with "- "several people or departments. null when unclear.\n"+ "- mode: 'team' when the text names colleagues at the user's own company "+ "(people with roles there, a team or a department list). 'solo' when it "+ "is only the user, or the only people named are outside their team (a "+ "boss they report to, clients, partners, advisors). null when unclear.\n""- role_kind: 'owner' (their own business), 'in_house' (leads a function ""or team in an organisation they do not own), 'independent' (advises or ""leads for clients), 'other', or null.\n"diff --git a/packages/core/tests/unit/test_onboarding_commit.py b/packages/core/tests/unit/test_onboarding_commit.pyindex 7ef707ef..db0eb183 100644--- a/packages/core/tests/unit/test_onboarding_commit.py+++ b/packages/core/tests/unit/test_onboarding_commit.py@@ -25,11 +25,13 @@ from openexecutive.onboarding.commit import (derive_org_structure,link_owner_email,reconcile_onboarding_departments,+ save_onboarding_goals,save_onboarding_people,)from openexecutive.onboarding.interview import (CompanyDraft,DepartmentDraft,+ GoalDraft,PersonDraft,)from openexecutive.people import store as people_store@@ -741,3 +743,79 @@ def test_adopt_never_renames_onto_an_existing_name(people_db: Path) -> None:def test_adopt_is_a_no_op_when_the_name_already_matches(people_db: Path) -> None:people_store.upsert_person(full_name="Ada Lovelace", email="[email protected]", is_principal=True)assert adopt_owner_row("[email protected]", "ada lovelace", "[email protected]") is False+++# ── goals ────────────────────────────────────────────────────────────────────+++def test_goals_are_saved_with_owners_and_no_department(empty_dept_db: Path) -> None:+ created = save_onboarding_goals(+ [+ GoalDraft(title="Launch the hosted version", target="Live by June"),+ GoalDraft(title="Ship the mobile apps", owner_name="dan lee"),+ GoalDraft(title=" "),+ ],+ {"Rufus Ames": 1, "Dan Lee": 2},+ principal_id=1,+ )+ assert created == 2+ goals = {g.key_result: g for g in dept_store.list_all_goals()}+ assert set(goals) == {"Launch the hosted version", "Ship the mobile apps"}+ launch = goals["Launch the hosted version"]+ assert launch.department_slug is None+ assert launch.owner_person_id == 1+ assert launch.target == "Live by June"+ assert launch.period_type == "year"+ assert goals["Ship the mobile apps"].owner_person_id == 2+ # No department was drafted, so none exists.+ assert dept_store.list_departments() == []+++def test_a_goal_lands_in_the_drafted_department(empty_dept_db: Path) -> None:+ reconcile_onboarding_departments([DepartmentDraft(title="Engineering")], {})+ save_onboarding_goals(+ [GoalDraft(title="Ship v2", department_title="engineering")], {}, principal_id=7+ )+ (goal,) = dept_store.list_all_goals()+ assert goal.department_slug == "engineering"+ # The principal, not the department's head: the review showed "you".+ assert goal.owner_person_id == 7+++def test_rerunning_setup_does_not_duplicate_goals(empty_dept_db: Path) -> None:+ drafts = [GoalDraft(title="Launch the hosted version")]+ assert save_onboarding_goals(drafts, {}, principal_id=1) == 1+ again = [GoalDraft(title="launch the hosted version "), GoalDraft(title="Win customers")]+ assert save_onboarding_goals(again, {}, principal_id=1) == 1+ assert sorted(g.key_result for g in dept_store.list_all_goals()) == [+ "Launch the hosted version",+ "Win customers",+ ]+++def test_last_years_goal_does_not_block_this_years(empty_dept_db: Path) -> None:+ dept_store.insert_goal(+ None, period_type="year", period_value="2001", key_result="Launch", target=""+ )+ assert save_onboarding_goals([GoalDraft(title="launch")], {}, principal_id=1) == 1+ assert sorted(g.period_value for g in dept_store.list_all_goals())[0] == "2001"+ assert len(dept_store.list_all_goals()) == 2+++def test_goal_failure_is_swallowed(monkeypatch: pytest.MonkeyPatch) -> None:+ def boom(*a: object, **k: object) -> None:+ raise sqlite3.OperationalError("disk I/O error")++ monkeypatch.setattr(dept_store, "initialize_db", boom)+ assert save_onboarding_goals([GoalDraft(title="Launch")], {}, principal_id=1) == 0+++def test_goals_become_this_years_priorities() -> None:+ d = _draft()+ goals = [GoalDraft(title="Launch"), GoalDraft(title=" Win customers ")]+ out = derive_org_structure(d.profile, d.people, d.departments, goals)+ assert out.strategic_priorities.current_year == ["Launch", "Win customers"]+ # No goals leaves the drafted priorities as they were.+ d.profile.strategic_priorities.current_year = ["Grow"]+ kept = derive_org_structure(d.profile, d.people, d.departments)+ assert kept.strategic_priorities.current_year == ["Grow"]diff --git a/packages/core/tests/unit/test_onboarding_interview.py b/packages/core/tests/unit/test_onboarding_interview.pyindex 7cec94fa..f79fe858 100644--- a/packages/core/tests/unit/test_onboarding_interview.py+++ b/packages/core/tests/unit/test_onboarding_interview.py@@ -380,9 +380,9 @@ async def test_solo_hint_follows_the_role_kind(def test_solo_hints_are_role_neutral_where_they_should_be() -> None:- """Only the owner hint asks about cash and runway; the neutral and the- in-house hints ask about the role; every hint keeps the one-person,- no-departments draft rule."""+ """No hint asks about cash or runway (setup never asks about finances);+ the neutral and the in-house hints ask about the role; every hint keeps+ the one-person, no-departments draft rule."""neutral = iv.SOLO_HINT.lower()assert "run the business alone" not in neutralassert "cash" not in neutral and "runway" not in neutral@@ -390,8 +390,9 @@ def test_solo_hints_are_role_neutral_where_they_should_be() -> None:assert word in neutral, wordowner = iv.SOLO_ROLE_HINTS["owner"].lower()- for word in ("offer", "customers", "pricing", "cash and runway", "top goals"):+ for word in ("offer", "customers", "pricing", "top goals"):assert word in owner, word+ assert "cash" not in owner and "runway" not in ownerin_house = iv.SOLO_ROLE_HINTS["in_house"].lower()assert "employer" in in_house@@ -812,3 +813,79 @@ async def test_prose_with_a_mid_text_question_mark_is_not_shown(_install(monkeypatch, provider)with pytest.raises(iv.InterviewError):await iv.advance(_opening(), questions_asked=0)+++# ── quick questions and goals ────────────────────────────────────────────────+++async def test_a_question_carries_its_suggested_answers(monkeypatch: pytest.MonkeyPatch) -> None:+ provider = _ScriptedProvider(+ [+ _tool_response(+ iv.ASK_TOOL_NAME,+ {"question": "Who is it for?", "options": ["Small teams", "Enterprises"]},+ )+ ]+ )+ _install(monkeypatch, provider)+ result = await iv.advance(_opening())+ assert isinstance(result, iv.Question)+ assert result.options == ["Small teams", "Enterprises"]+ ask = next(t for t in provider.calls[0]["tools"] if t["name"] == iv.ASK_TOOL_NAME)+ assert "options" in ask["input_schema"]["properties"]+++def test_bad_options_cost_the_buttons_not_the_question() -> None:+ long = "x" * (iv.MAX_OPTION_CHARS + 1)+ q = iv.Question.model_validate(+ {"question": "Q?", "options": ["A", " ", "A", long, 3, "B", "C", "D", "E"]}+ )+ assert q.options == ["A", "B", "C", "D"]+ assert iv.Question.model_validate({"question": "Q?", "options": "A"}).options == []+ assert iv.Question.model_validate({"question": "Q?", "options": None}).options == []+++def test_setup_asks_at_most_four_questions() -> None:+ assert iv.MAX_QUESTIONS == 4+++def test_validate_draft_accepts_goals_with_known_owners_and_departments() -> None:+ draft = _draft(+ goals=[+ {"title": "Launch same-day delivery", "owner_name": "sam okafor"},+ {"title": "Cut returns", "department_title": "operations"},+ {"title": "Hire a CFO", "owner_name": None, "target": None},+ ]+ )+ assert iv.validate_draft(draft) == []+ assert draft.goals[2].owner_name == "" and draft.goals[2].target == ""+++def test_validate_draft_rejects_a_goal_owned_by_a_stranger() -> None:+ draft = _draft(goals=[{"title": "Launch", "owner_name": "Pat Doe"}])+ errors = iv.validate_draft(draft)+ assert any("owner_name 'Pat Doe'" in e for e in _details(errors))+ assert all("Pat Doe" not in e for e in _safes(errors))+++def test_validate_draft_rejects_a_goal_in_an_undrafted_department() -> None:+ draft = _draft(goals=[{"title": "Launch", "department_title": "Marketing"}])+ assert any("department_title 'Marketing'" in e for e in _details(iv.validate_draft(draft)))+++def test_validate_draft_bounds_goals() -> None:+ blank = _draft(goals=[{"title": " "}])+ assert "Every goal needs a name." in _safes(iv.validate_draft(blank))+ many = _draft(goals=[{"title": f"Goal {i}"} for i in range(iv.MAX_GOALS + 1)])+ assert any("lot of goals" in e for e in _safes(iv.validate_draft(many)))+ long = _draft(goals=[{"title": "x" * (iv.MAX_GOAL_CHARS + 1)}])+ assert any("goal is too long" in e for e in _safes(iv.validate_draft(long)))+++def test_the_prompt_never_asks_about_departments_or_finances() -> None:+ text = ONBOARDING_INTERVIEWER_SYSTEM+ assert "Never ask about departments" in text+ assert "Financials: never ask" in text+ assert "ONLY when the user's own words or files already list" in text+ assert "never use field names" in textdiff --git a/packages/core/tests/unit/test_onboarding_interview_route.py b/packages/core/tests/unit/test_onboarding_interview_route.pyindex cda6a31f..b2b10312 100644--- a/packages/core/tests/unit/test_onboarding_interview_route.py+++ b/packages/core/tests/unit/test_onboarding_interview_route.py@@ -110,10 +110,16 @@ def client("openexecutive.onboarding.commit.reconcile_onboarding_departments",lambda d, ids: {"updated": 0, "created": 0},)+ saved_goals: list[Any] = []+ monkeypatch.setattr(+ "openexecutive.onboarding.commit.save_onboarding_goals",+ lambda goals, ids, principal_id: saved_goals.append((goals, principal_id)) or len(goals),+ )app = FastAPI()app.include_router(route.router)c = TestClient(app)c.fired = fired # type: ignore[attr-defined]+ c.saved_goals = saved_goals # type: ignore[attr-defined]return c@@ -1491,3 +1497,62 @@ def test_a_failed_revise_never_rolls_back_a_turn_added_meanwhile(assert resp.status_code == 502assert held.transcript[-1].text == "added meanwhile"assert held.transcript[0].text == "We are Northwind."+++# ── quick questions and goals ────────────────────────────────────────────────+++def test_a_question_comes_back_with_its_suggested_answers(+ client: TestClient, seeded: list[Any]+) -> None:+ seeded.append(iv.Question(question="Who is it for?", options=["Small teams", "Enterprises"]))+ resp = client.post("/onboard/interview/start", data={"description": "We sell tools."})+ assert resp.json()["question_options"] == ["Small teams", "Enterprises"]+ sid = resp.json()["session_id"]+ # Kept across a refresh, and gone once a draft replaces the question.+ assert client.get(f"/onboard/interview/{sid}").json()["question_options"] == [+ "Small teams",+ "Enterprises",+ ]+ seeded.append(_draft())+ drafted = client.post("/onboard/interview/message", json={"session_id": sid, "message": "Small teams"})+ assert drafted.json()["question_options"] == []+++def test_a_draft_shows_its_goals(client: TestClient, seeded: list[Any]) -> None:+ seeded.append(_draft(goals=[{"title": "Launch same-day delivery", "owner_name": "Sam Okafor"}]))+ resp = client.post("/onboard/interview/start", data={"description": "We sell tools."})+ assert resp.json()["draft_goals"] == [+ {+ "title": "Launch same-day delivery",+ "target": "",+ "owner_name": "Sam Okafor",+ "department_title": "",+ }+ ]+++def test_commit_saves_the_goals(client: TestClient, seeded: list[Any], profile_path: Path) -> None:+ seeded.append(_draft())+ sid = _start(client)+ body = _commit_body(sid)+ body["goals"] = [{"title": "Launch same-day delivery"}, {"title": "Cut returns"}]+ assert client.post("/onboard/interview/commit", json=body).status_code == 200+ ((goals, _principal),) = client.saved_goals # type: ignore[attr-defined]+ assert [g.title for g in goals] == ["Launch same-day delivery", "Cut returns"]+ saved = CompanyProfile.load_from_yaml(profile_path)+ assert saved.strategic_priorities.current_year == ["Launch same-day delivery", "Cut returns"]+++def test_commit_rejects_a_goal_owned_by_a_stranger_without_echoing(+ client: TestClient, seeded: list[Any], profile_path: Path+) -> None:+ seeded.append(_draft())+ sid = _start(client)+ body = _commit_body(sid)+ body["goals"] = [{"title": "Launch", "owner_name": "Zed Secretname"}]+ resp = client.post("/onboard/interview/commit", json=body)+ assert resp.status_code == 422+ assert "Secretname" not in resp.text+ assert not profile_path.exists()+ assert client.saved_goals == [] # type: ignore[attr-defined]diff --git a/packages/core/tests/unit/test_onboarding_interview_store.py b/packages/core/tests/unit/test_onboarding_interview_store.pyindex 46751a6d..c06dfc89 100644--- a/packages/core/tests/unit/test_onboarding_interview_store.py+++ b/packages/core/tests/unit/test_onboarding_interview_store.py@@ -32,6 +32,7 @@ def _session() -> InterviewSession:questions_asked=1,draft=draft,last_hint="A name.",+ last_options=["Sam Okafor", "Nobody yet"],)@@ -42,6 +43,7 @@ def test_round_trip_keeps_every_field() -> None:assert [t.text for t in back.transcript] == ["We sell tools.", "Who runs ops?"]assert back.questions_asked == 1assert back.last_hint == "A name."+ assert back.last_options == ["Sam Okafor", "Nobody yet"]assert back.saved is Falseassert back.draft is not Noneassert back.draft.profile.name == "Northwind Tools"diff --git a/packages/ui/AGENTS.md b/packages/ui/AGENTS.mdindex 643577df..af3b7c84 100644--- a/packages/ui/AGENTS.md+++ b/packages/ui/AGENTS.md@@ -1,6 +1,6 @@<!-- BEGIN:nextjs-agent-rules -->-# This is NOT the Next.js you know+## This is NOT the Next.js you knowThis version has breaking changes — APIs, conventions, and file structure may all differ from your training data. Read the relevant guide in `node_modules/next/dist/docs/` (resolved from this file's directory; in monorepos the `next` package may not be visible from the repo root) before writing any code. Heed deprecation notices.diff --git a/packages/ui/next-env.d.ts b/packages/ui/next-env.d.tsindex ce4e94a6..a419cbe4 100644--- a/packages/ui/next-env.d.ts+++ b/packages/ui/next-env.d.ts@@ -1,7 +1,7 @@/// <reference types="next" />/// <reference types="next/image-types/global" />-import "./.next/types/routes.d.ts";-import "./.next/types/root-params.d.ts";+import "./.next/dev/types/routes.d.ts";+import "./.next/dev/types/root-params.d.ts";// NOTE: This file should not be edited// see https://nextjs.org/docs/app/api-reference/config/typescript for more information.diff --git a/packages/ui/scripts/onboardSetup.test.mjs b/packages/ui/scripts/onboardSetup.test.mjsnew file mode 100644index 00000000..46b127db--- /dev/null+++ b/packages/ui/scripts/onboardSetup.test.mjs@@ -0,0 +1,76 @@+import assert from "node:assert/strict";+import test from "node:test";+import {+ afterSaveSteps,+ checkNotes,+ goalsToSave,+ inferredMode,+ modeLine,+ understoodRole,+} from "../src/lib/onboardSetup.ts";++const profile = (over = {}) => ({+ competitive_landscape: { primary_competitors: [], competitive_advantages: [] },+ vendors: [],+ tickers: [],+ ...over,+});+const goal = (title, owner_name = "", department_title = "") => ({ title, target: "", owner_name, department_title });++test("setup picks Personal unless the text says Team", () => {+ assert.equal(inferredMode(null), "solo");+ assert.equal(inferredMode({ mode: null }), "solo");+ assert.equal(inferredMode({ mode: "team" }), "team");+});++test("only the role fields the text gave are saved", () => {+ assert.deepEqual(understoodRole(null), {});+ assert.deepEqual(understoodRole({ role_kind: "owner", role_title: "", reports_to: null }), { role_kind: "owner" });+ assert.deepEqual(understoodRole({ role_kind: null, role_title: "COO", reports_to: "CEO" }), {+ role_title: "COO",+ reports_to: "CEO",+ });+});++test("the mode line names the team, up to four", () => {+ assert.equal(modeLine("solo", ["Rufus", "Dan"]), "Set up just for you");+ assert.equal(modeLine("team", []), "Set up for your team");+ assert.equal(modeLine("team", ["Rufus", " ", "Dan"]), "Set up for your team: Rufus, Dan");+ assert.equal(+ modeLine("team", ["A", "B", "C", "D", "E", "F"]),+ "Set up for your team: A, B, C, D and 2 more",+ );+});++test("notes that speak in field names never reach the review", () => {+ assert.deepEqual(+ checkNotes(["annual_revenue_arr not stated", " ", "I guessed your competitors from your site."]),+ ["I guessed your competitors from your site."],+ );+});++test("after save always researches, and mentions goals and watched services when there are some", () => {+ assert.deepEqual(afterSaveSteps(profile(), []), [+ "Research your competitors and suggest a shortlist in your brief",+ ]);+ const steps = afterSaveSteps(+ profile({ competitive_landscape: { primary_competitors: ["Acme"] }, vendors: ["Fly.io"], tickers: ["NET"] }),+ [goal("Launch")],+ );+ assert.equal(steps.length, 3);+ assert.match(steps[0], /market and competitors/);+ assert.match(steps[1], /every week/);+ assert.equal(steps[2], "Watch the status pages and news for Fly.io, NET");+});++test("goals drop blanks and owners or departments no longer on the review", () => {+ const out = goalsToSave(+ [goal(" Launch ", "dan", "Engineering"), goal(" "), goal("Win customers", "Gone", "Sales")],+ ["Rufus", "Dan"],+ ["engineering"],+ );+ assert.deepEqual(out, [+ { title: "Launch", target: "", owner_name: "dan", department_title: "Engineering" },+ { title: "Win customers", target: "", owner_name: "", department_title: "" },+ ]);+});diff --git a/packages/ui/src/app/onboard/page.tsx b/packages/ui/src/app/onboard/page.tsxindex a7dd4c5b..bd66f67b 100644--- a/packages/ui/src/app/onboard/page.tsx+++ b/packages/ui/src/app/onboard/page.tsx@@ -8,9 +8,7 @@ import OnboardConversation, {} from "@/components/onboard/OnboardConversation";import OnboardAnswers from "@/components/onboard/OnboardAnswers";import OnboardDescribe from "@/components/onboard/OnboardDescribe";-import OnboardDone from "@/components/onboard/OnboardDone";import OnboardDraftReview from "@/components/onboard/OnboardDraftReview";-import VoicePicker from "@/components/executive/VoicePicker";import { useWorkspace } from "@/components/workspace/WorkspaceContext";import {discardOnboardInterview,@@ -24,13 +22,13 @@ import { clearSetupMemory, recall, remember, STORAGE_KEYS } from "@/lib/firstRun// (see AppShell.tsx EXEMPT_PREFIXES) so it owns the whole viewport.//// First run starts with the free-text box: describe your work (and attach a-// one-pager if you like). The Executive reads it, shows what it understood —-// personal or team, role, company, focus — for a quick confirm, saves the-// workspace mode and role, and then asks only about what is still missing-// (up to five questions, each skippable). Then the drafted profile is-// reviewed and, last, how the Executive should sound (skippable; changeable-// in Settings). A re-run — a company profile already exists — skips the-// first two: they are changed in Settings.+// one-pager if you like). The Executive reads it and saves Personal or Team+// and the role without asking (Personal when it can't tell), then asks a few+// quick questions about what is still missing (up to four, each with answers+// to tap, each skippable). Then the drafted profile, people and goals are+// reviewed, with Personal or Team on one line and a Change link, and Save+// goes to Home, where the getting-started checklist carries on. A re-run — a+// company profile already exists — skips the box and keeps the saved mode.//// The original step-by-step wizard stays reachable at /onboard?mode=form — it// needs no API key beyond the profile save, so it is also the fallback when@@ -42,8 +40,7 @@ import { clearSetupMemory, recall, remember, STORAGE_KEYS } from "@/lib/firstRun// A setup in progress survives a refresh, a closed tab and a restart of the// API: the interview is kept on the server, and this browser remembers which// one is theirs (lib/firstRun.ts). Opening /onboard with one remembered-// picks it up at the same question or draft. After the save, a last screen-// shows what was kept and the two accounts worth connecting.+// picks it up at the same question or draft.const FOR_PARAM: Record<WorkspaceMode, string> = { solo: "me", team: "team" };@@ -66,8 +63,6 @@ function OnboardFlow() {const [hasProfile, setHasProfile] = useState<boolean | null>(null);// Set once the describe-first step is done, in place of the `?for=` param.const [chosenHere, setChosenHere] = useState<WorkspaceMode | null>(null);- const [askVoice, setAskVoice] = useState(false);- const [showDone, setShowDone] = useState(false);// Resuming: the remembered session is being fetched.const [resuming, setResuming] = useState(() => recall(STORAGE_KEYS.session) !== null);const [resumed, setResumed] = useState(false);@@ -150,11 +145,11 @@ function OnboardFlow() {</div>);+ // Home, where the getting-started checklist lives. The tone of voice and+ // everything else stay in Settings.function finish() {clearSetupMemory();- // Only a first setup asks the voice; a re-run keeps what it has.- if (chosenHere || resumed) setAskVoice(true);- else router.push("/");+ router.push("/");}function later() {@@ -165,10 +160,6 @@ function OnboardFlow() {if (hasProfile === null || resuming) return null;- if (showDone) return <OnboardDone solo={mode === "solo"} onDone={() => router.push("/")} />;-- if (askVoice) return <VoiceStep onDone={() => setShowDone(true)} />;-if (!hasProfile && !chosenFor && !chosenHere && !resumed) {return (<OnboardDescribe@@ -241,10 +232,15 @@ function OnboardFlow() {return (<div className="max-w-2xl mx-auto px-4 sm:px-6 py-16 w-full"><div className="mb-8">- <h1 className="text-xl font-semibold text-fg">Set up your Executive</h1>+ <h1 className="text-xl font-semibold text-fg">+ {conversationTurn ? "A few quick questions" : "Set up your Executive"}+ </h1><p className="text-sm text-fg-muted mt-1">- A few minutes now, and every answer you get afterwards is grounded in- your {mode === "solo" ? "work" : "company"} rather than a generic one.+ {conversationTurn+ ? "Tap an answer, type your own, or skip. Every question is optional."+ : `A few minutes now, and every answer you get afterwards is grounded in your ${+ mode === "solo" ? "work" : "company"+ } rather than a generic one.`}</p></div>@@ -275,23 +271,6 @@ function OnboardFlow() {);}-// How should the Executive sound? Saved on the Executive (PATCH-// /agents/executive); skipping keeps Direct. Settings → Executive changes it.-function VoiceStep({ onDone }: { onDone: () => void }) {- return (- <div className="max-w-3xl mx-auto px-4 sm:px-6 py-16 w-full">- <h1 className="text-xl font-semibold text-fg">How should your Executive sound?</h1>- <p className="text-sm text-fg-muted mt-1">- The advice is the same; this sets the tone it comes in. You can change it later in- Settings.- </p>- <div className="mt-8">- <VoicePicker variant="step" onDone={onDone} />- </div>- </div>- );-}-export default function OnboardPage() {return (<div className="flex flex-col h-full bg-surface">diff --git a/packages/ui/src/components/onboard/ExecEmailRow.tsx b/packages/ui/src/components/onboard/ExecEmailRow.tsxdeleted file mode 100644index f9647833..00000000--- a/packages/ui/src/components/onboard/ExecEmailRow.tsx+++ /dev/null@@ -1,24 +0,0 @@-"use client";--import Link from "next/link";-import { FIRST_RUN_LINKS } from "@/lib/firstRun";--// The last setup screen's line about the Executive's own email: an account-// just for it, which it reads, sends and keeps a calendar as. Its own file so-// a deployment that gives every Executive an address can say that instead.-export default function ExecEmailRow() {- return (- <li className="flex items-baseline gap-3 border-b border-line py-3 last:border-0">- <span aria-hidden className="w-5 flex-shrink-0 text-center">- ✉- </span>- <p className="min-w-0 text-[15px] text-fg-muted leading-snug">- <span className="font-semibold text-fg">Give it its own email.</span> A Google or Microsoft account just- for it, so it can send, keep a calendar and email your morning brief.{" "}- <Link href={FIRST_RUN_LINKS.execMailbox} className="font-semibold text-accent hover:underline">- Connect its account- </Link>- </p>- </li>- );-}diff --git a/packages/ui/src/components/onboard/OnboardConversation.tsx b/packages/ui/src/components/onboard/OnboardConversation.tsxindex d19da930..690b371b 100644--- a/packages/ui/src/components/onboard/OnboardConversation.tsx+++ b/packages/ui/src/components/onboard/OnboardConversation.tsx@@ -47,6 +47,9 @@ function flavourFor(solo: boolean, roleKind: RoleKind | null | undefined): Flavoreturn "solo";}+// What Skip sends. The interviewer moves on and does not ask it again.+const SKIP_ANSWER = "Skip this one.";+export interface Bubble {role: "user" | "assistant";text: string;@@ -96,8 +99,9 @@ export default function OnboardConversation({]);}- async function send() {- const text = input.trim();+ // A tapped suggestion or Skip sends its own text; otherwise the box's.+ async function send(tapped?: string) {+ const text = (tapped ?? input).trim();if ((!text && files.length === 0) || busy) return;setBusy(true);setError(null);@@ -141,6 +145,29 @@ export default function OnboardConversation({{turn?.question_hint && (<p className="text-xs text-fg-muted mt-2">{turn.question_hint}</p>)}+ {started && headline && (+ <div className="flex flex-wrap gap-2 mt-3">+ {(turn?.question_options ?? []).map((option) => (+ <button+ key={option}+ type="button"+ onClick={() => void send(option)}+ disabled={busy}+ className="min-h-9 px-3.5 py-1.5 rounded-full border border-line-strong text-sm text-fg text-left hover:border-indigo-500/60 hover:bg-indigo-500/10 disabled:opacity-40 transition-colors cursor-pointer"+ >+ {option}+ </button>+ ))}+ <button+ type="button"+ onClick={() => void send(SKIP_ANSWER)}+ disabled={busy}+ className="min-h-9 px-3 py-1.5 text-sm text-fg-muted hover:text-fg disabled:opacity-40 transition-colors cursor-pointer"+ >+ Skip+ </button>+ </div>+ )}</div></div>@@ -160,7 +187,7 @@ export default function OnboardConversation({autoFocusplaceholder={started- ? "Your answer… (Enter to send, Shift+Enter for a new line)"+ ? "Or type your own answer… (Enter to send, Shift+Enter for a new line)": OPENING_EXAMPLE[flavour]}className="w-full rounded-lg border border-line-strong bg-surface-overlay px-3 py-2 text-sm text-fg placeholder-fg-subtle focus:outline-none focus:ring-2 focus:ring-indigo-500/50 focus:border-indigo-500/50 resize-none transition-colors disabled:opacity-50"@@ -204,12 +231,12 @@ export default function OnboardConversation({disabled={busy}className="text-xs text-fg-muted hover:text-fg disabled:opacity-40 transition-colors">- Skip ahead — draft my profile now+ Draft it now</button>)}{started && (<span className="text-xs text-fg-subtle ml-auto">- {turn!.questions_asked} of {turn!.max_questions} questions+ {turn!.questions_asked} of up to {turn!.max_questions}</span>)}</div>diff --git a/packages/ui/src/components/onboard/OnboardDescribe.tsx b/packages/ui/src/components/onboard/OnboardDescribe.tsxindex 30ef4aba..eb8ca111 100644--- a/packages/ui/src/components/onboard/OnboardDescribe.tsx+++ b/packages/ui/src/components/onboard/OnboardDescribe.tsx@@ -1,9 +1,9 @@"use client";import { useEffect, useState } from "react";-import RoleFields from "@/components/workspace/RoleFields";import { useWorkspace } from "@/components/workspace/WorkspaceContext";import { recall, remember, STORAGE_KEYS } from "@/lib/firstRun";+import { inferredMode, understoodRole } from "@/lib/onboardSetup";import {startOnboardInterview,understandOnboarding,@@ -12,29 +12,19 @@ import {type OnboardUnderstanding,type WorkspaceMode,} from "@/lib/api";-import {- EMPTY_ROLE_FORM,- roleFormErrors,- roleUpdate,- type RoleForm,-} from "@/lib/principalRole";-// First run: describe your work, then confirm what was understood.+// First run: describe your work in one box (and optional files), then the+// interview starts. Nothing is asked in between.//-// Step 1 is one box (and optional files). Step 2 shows what the text already-// said — personal or team, role, company, focus — each editable, and asks only-// for what it could not tell (personal vs team is the one required answer).-// Continue saves the workspace mode and role, then opens the interview with-// the same text, which asks about whatever is still missing. Nothing is asked-// twice: the role the form shows is what the interview is told.+// Continue reads the text once (POST /onboard/interview/understand) and saves+// what it can tell straight away: Personal or Team (Personal when it can't+// tell, since nothing is lost and switching is easy), the browser's time zone+// and, for Personal, the role the text gave. The review screen shows the+// choice on one line with a Change link, and Settings keeps the switch. Then+// the interview opens with the same text and asks only about what is missing.const MAX_FILES = 8;-const MODES: { mode: WorkspaceMode; title: string; lead: string }[] = [- { mode: "solo", title: "Personal", lead: "Your own executive: your goals, your inbox, your mornings." },- { mode: "team", title: "Team", lead: "An executive for the whole company, with a lead for each department." },-];-const EXAMPLE ="e.g. I'm Head of Customer Success at a 300-person software company and report to the COO. My team of twelve looks after renewals. This year I want net revenue retention above 110% and a cleaner handoff from sales.";@@ -52,16 +42,6 @@ export interface DescribeResult {turns: { role: "user" | "assistant"; text: string }[];}-function Steps({ at }: { at: 1 | 2 }) {- return (- <div className="flex gap-1.5 mb-6" aria-hidden>- {[1, 2, 3].map((i) => (- <i key={i} className={`h-1 flex-1 rounded-full ${i <= at ? "bg-indigo-500" : "bg-line"}`} />- ))}- </div>- );-}-interface DescribeProps {onReady: (r: DescribeResult) => void;/** "I'll do this later": Home, with the checklist keeping setup one tap away. */@@ -69,18 +49,11 @@ interface DescribeProps {}export default function OnboardDescribe({ onReady, onLater }: DescribeProps) {- const { role, refresh } = useWorkspace();+ const { refresh } = useWorkspace();// What was typed survives a refresh until Continue starts the interview.const [text, setText] = useState(() => recall(STORAGE_KEYS.description) ?? "");const [files, setFiles] = useState<File[]>([]);- const [step, setStep] = useState<"describe" | "confirm">("describe");- const [reading, setReading] = useState(false);- const [note, setNote] = useState<string | null>(null);- const [understood, setUnderstood] = useState<OnboardUnderstanding | null>(null);- const [mode, setMode] = useState<WorkspaceMode | null>(null);- const [form, setForm] = useState<RoleForm>(EMPTY_ROLE_FORM);- const [editing, setEditing] = useState(false);- const [saving, setSaving] = useState(false);+ const [busy, setBusy] = useState(false);const [error, setError] = useState<string | null>(null);const hasInput = text.trim().length > 0 || files.length > 0;@@ -89,243 +62,117 @@ export default function OnboardDescribe({ onReady, onLater }: DescribeProps) {remember(STORAGE_KEYS.description, text.trim() ? text : null);}, [text]);- function toConfirm(u: OnboardUnderstanding | null) {- setUnderstood(u);- setMode(u?.mode ?? null);- setForm({- ...EMPTY_ROLE_FORM,- role_kind: u?.role_kind ?? null,- role_title: u?.role_title ?? "",- reports_to: u?.reports_to ?? "",- });- setStep("confirm");- }-- async function read() {- if (!hasInput || reading) return;- setReading(true);- setNote(null);+ async function saveWorkspace(u: OnboardUnderstanding | null): Promise<WorkspaceMode> {+ const mode = inferredMode(u);+ const timezone = browserTimeZone();try {- toConfirm(await understandOnboarding(text.trim(), files));- } catch {- // Not fatal: the text is kept and the person just picks below.- setNote("I couldn't read that automatically, so pick below. Your text is kept.");- toConfirm(null);- } finally {- setReading(false);+ await updateWorkspace(timezone ? { mode, timezone } : { mode });+ } catch (err) {+ // A zone the server does not know must not block setup.+ if (!timezone || (err instanceof Error && /principal/i.test(err.message))) throw err;+ await updateWorkspace({ mode });+ }+ const role = understoodRole(u);+ if (mode === "solo" && Object.keys(role).length > 0) {+ // Best effort: the interview asks about the role anyway.+ await updateWorkspace(role).catch(() => undefined);}+ await refresh();+ return mode;}- const problems = roleFormErrors(form);-- async function carryOn() {- if (!mode || saving || problems.length > 0) return;- setSaving(true);+ async function begin(read: boolean) {+ if (busy || (read && !hasInput)) return;+ setBusy(true);setError(null);try {- const timezone = browserTimeZone();- try {- await updateWorkspace(timezone ? { mode, timezone } : { mode });- } catch (err) {- // A zone the server does not know must not block setup.- if (!timezone || (err instanceof Error && /principal/i.test(err.message))) throw err;- await updateWorkspace({ mode });- }- if (mode === "solo") {- const update = roleUpdate(form, role);- if (Object.keys(update).length > 0) await updateWorkspace(update);+ let understood: OnboardUnderstanding | null = null;+ if (read) {+ // Not fatal: without it setup starts Personal and the review says so.+ understood = await understandOnboarding(text.trim(), files).catch(() => null);}- await refresh();- const description = text.trim();- const turn = await startOnboardInterview(description, files);+ const mode = await saveWorkspace(understood);+ const description = read ? text.trim() : "";+ const turn = await startOnboardInterview(description, read ? files : []);const turns: DescribeResult["turns"] = [];// Files alone still make the opening answer (the server's first turn),// so the answers line up with the server's when one is edited.- if (description || files.length > 0) turns.push({ role: "user", text: description });+ if (description || (read && files.length > 0)) turns.push({ role: "user", text: description });if (turn.phase === "question" && turn.question) {turns.push({ role: "assistant", text: turn.question });}onReady({ mode, turn, turns });} catch (err) {- setError(err instanceof Error ? err.message : "Could not save that.");- setSaving(false);+ setError(err instanceof Error ? err.message : "Could not start setup.");+ setBusy(false);}}- if (step === "describe") {- return (- <div className="max-w-2xl mx-auto px-4 sm:px-6 py-12 sm:py-16 w-full">- <Steps at={1} />- <h1 className="text-xl font-semibold text-fg">Set up your Executive</h1>- <p className="text-sm text-fg-muted mt-1">- Tell me about your work in your own words. I'll work out the rest and only ask- about what's missing.- </p>- <textarea- value={text}- onChange={(e) => setText(e.target.value)}- rows={7}- autoFocus- disabled={reading}- placeholder={EXAMPLE}- aria-label="Describe your work"- className="mt-6 w-full rounded-lg border border-line-strong bg-surface-overlay px-3 py-2 text-sm text-fg placeholder-fg-subtle focus:outline-none focus:ring-2 focus:ring-indigo-500/50 focus:border-indigo-500/50 resize-none transition-colors disabled:opacity-50"- />- <div className="mt-2 flex items-center gap-3">- <label className="text-xs text-indigo-400 hover:text-indigo-300 cursor-pointer transition-colors">- + Attach a one-pager, deck or CV (optional)- <input- type="file"- multiple- className="hidden"- accept=".pdf,.docx,.doc,.xlsx,.xlsm,.csv,.md,.txt"- onChange={(e) => setFiles(Array.from(e.target.files ?? []).slice(0, MAX_FILES))}- />- </label>- {files.length > 0 && (- <p className="text-xs text-fg-muted truncate">{files.map((f) => f.name).join(", ")}</p>- )}- </div>- <div className="mt-6 flex flex-wrap items-center gap-x-4 gap-y-2">- <button- type="button"- onClick={() => void read()}- disabled={!hasInput || reading}- className="px-5 py-2.5 rounded-lg text-sm font-medium bg-indigo-500 hover:bg-indigo-600 text-white transition-colors cursor-pointer disabled:opacity-50 disabled:cursor-not-allowed"- >- {reading ? "Reading…" : "Continue"}- </button>- <button- type="button"- onClick={() => toConfirm(null)}- disabled={reading}- className="text-sm text-fg-muted hover:text-fg transition-colors cursor-pointer disabled:opacity-50"- >- Skip, I'll answer questions instead- </button>- </div>- {onLater && (- <p className="mt-10 text-sm text-fg-subtle">- Not now?{" "}- <button- type="button"- onClick={onLater}- disabled={reading}- className="font-semibold text-accent hover:underline cursor-pointer disabled:opacity-50"- >- I'll do this later- </button>- . Home keeps setup one tap away.- </p>- )}- </div>- );- }-- const facts: [string, string | null][] = [- ["Company", understood?.company ?? null],- ["This year", understood?.focus ?? null],- ];- const shown = facts.filter(([, v]) => v);-return (<div className="max-w-2xl mx-auto px-4 sm:px-6 py-12 sm:py-16 w-full">- <Steps at={2} />- <h1 className="text-xl font-semibold text-fg">- {understood ? "Here’s what I got" : "Who is this for?"}- </h1>+ <h1 className="text-xl font-semibold text-fg">Set up your Executive</h1><p className="text-sm text-fg-muted mt-1">- {understood- ? "Fix anything that’s off. It all stays editable in Settings."- : "One tap. You can switch in Settings or add teammates anytime."}+ Tell me about your work in your own words. I'll work out the rest and ask a few+ quick questions about anything that's missing.</p>- {note && <p className="text-xs text-amber-400 mt-2">{note}</p>}-- <div role="radiogroup" aria-label="Set up for" className="mt-6 grid gap-3 sm:grid-cols-2">- {MODES.map((m) => {- const on = mode === m.mode;- return (- <button- key={m.mode}- type="button"- role="radio"- aria-checked={on}- disabled={saving}- onClick={() => setMode(m.mode)}- className={`text-left rounded-xl border p-4 transition-colors cursor-pointer focus:outline-none focus:ring-2 focus:ring-indigo-500/50 disabled:opacity-60 ${- on- ? "border-indigo-500/70 bg-indigo-500/10"- : "border-line bg-surface-elevated hover:border-line-strong"- }`}- >- <span className="block text-sm font-semibold text-fg">{m.title}</span>- <span className="block text-xs text-fg-muted mt-1 leading-snug">{m.lead}</span>- </button>- );- })}+ <textarea+ value={text}+ onChange={(e) => setText(e.target.value)}+ rows={7}+ autoFocus+ disabled={busy}+ placeholder={EXAMPLE}+ aria-label="Describe your work"+ className="mt-6 w-full rounded-lg border border-line-strong bg-surface-overlay px-3 py-2 text-sm text-fg placeholder-fg-subtle focus:outline-none focus:ring-2 focus:ring-indigo-500/50 focus:border-indigo-500/50 resize-none transition-colors disabled:opacity-50"+ />+ <div className="mt-2 flex items-center gap-3">+ <label className="text-xs text-indigo-400 hover:text-indigo-300 cursor-pointer transition-colors">+ + Attach a one-pager, deck or CV (optional)+ <input+ type="file"+ multiple+ className="hidden"+ accept=".pdf,.docx,.doc,.xlsx,.xlsm,.csv,.md,.txt"+ onChange={(e) => setFiles(Array.from(e.target.files ?? []).slice(0, MAX_FILES))}+ />+ </label>+ {files.length > 0 && (+ <p className="text-xs text-fg-muted truncate">{files.map((f) => f.name).join(", ")}</p>+ )}</div>-- {shown.length > 0 && (- <dl className="mt-4 rounded-xl border border-line bg-surface-elevated px-4 py-1">- {shown.map(([k, v]) => (- <div key={k} className="flex justify-between gap-4 py-2.5 text-sm border-b border-line last:border-0">- <dt className="text-fg-muted">{k}</dt>- <dd className="text-fg text-right">{v}</dd>- </div>- ))}- </dl>- )}-- {mode === "solo" && (- <div className="mt-4 rounded-xl border border-line bg-surface-elevated p-4">- {!editing && (form.role_kind || form.role_title || form.reports_to) ? (- <div className="flex items-start justify-between gap-4 text-sm">- <div className="text-fg">- {[form.role_title, form.reports_to && `reports to ${form.reports_to}`]- .filter(Boolean)- .join(" · ") || "Role set"}- </div>- <button- type="button"- onClick={() => setEditing(true)}- className="text-xs text-indigo-400 hover:text-indigo-300 cursor-pointer"- >- Edit- </button>- </div>- ) : (- <RoleFields value={form} onChange={setForm} disabled={saving} idPrefix="describe-role" />- )}- </div>- )}-- {(problems.length > 0 || error) && (- <div className="mt-3 space-y-1 text-xs text-red-400">- {problems.map((p) => (- <p key={p}>{p}</p>- ))}- {error && <p>{error}</p>}- </div>- )}-+ {error && <p className="mt-3 text-xs text-red-400">{error}</p>}<div className="mt-6 flex flex-wrap items-center gap-x-4 gap-y-2"><buttontype="button"- onClick={() => void carryOn()}- disabled={!mode || saving || problems.length > 0}+ onClick={() => void begin(true)}+ disabled={!hasInput || busy}className="px-5 py-2.5 rounded-lg text-sm font-medium bg-indigo-500 hover:bg-indigo-600 text-white transition-colors cursor-pointer disabled:opacity-50 disabled:cursor-not-allowed">- {saving ? "Saving…" : "Continue"}+ {busy ? "Reading…" : "Continue"}</button><buttontype="button"- onClick={() => setStep("describe")}- disabled={saving}+ onClick={() => void begin(false)}+ disabled={busy}className="text-sm text-fg-muted hover:text-fg transition-colors cursor-pointer disabled:opacity-50">- Back+ Skip, I'll answer questions instead</button></div>+ {onLater && (+ <p className="mt-10 text-sm text-fg-subtle">+ Not now?{" "}+ <button+ type="button"+ onClick={onLater}+ disabled={busy}+ className="font-semibold text-accent hover:underline cursor-pointer disabled:opacity-50"+ >+ I'll do this later+ </button>+ . Home keeps setup one tap away.+ </p>+ )}</div>);}diff --git a/packages/ui/src/components/onboard/OnboardDone.tsx b/packages/ui/src/components/onboard/OnboardDone.tsxdeleted file mode 100644index b82c483f..00000000--- a/packages/ui/src/components/onboard/OnboardDone.tsx+++ /dev/null@@ -1,142 +0,0 @@-"use client";--import Link from "next/link";-import { useEffect, useRef, useState } from "react";-import ExecEmailRow from "@/components/onboard/ExecEmailRow";-import { buttonClass } from "@/components/ui/Button";-import { getCompanyProfile, listPeople } from "@/lib/api";-import { FIRST_RUN_LINKS, OWN_MAILBOX_SENDS, STORAGE_KEYS, firstQuestion, remember } from "@/lib/firstRun";--interface Props {- /** Personal (one person) or a team; sets the wording. */- solo: boolean;- /** Opens Home. */- onDone: () => void;-}--// The last setup screen ends on a conversation: one question, built from the-// priorities setup just saved, which Home asks as soon as it opens (handed-// over in session storage, never the URL, so a link can't make someone ask-// something). Under it, what the Executive can already do, written as things-// to try: its own email, the brief, drafting replies as you. The connections-// are also rows on Home's getting-started checklist, so nothing is lost by-// asking first.-export default function OnboardDone({ solo, onDone }: Props) {- const [companyName, setCompanyName] = useState<string | null>(null);- const [peopleCount, setPeopleCount] = useState<number | null>(null);- const [question, setQuestion] = useState<string>(() => firstQuestion(null, solo));- // Set once they type: a profile that loads late must not overwrite them.- const edited = useRef(false);-- useEffect(() => {- let cancelled = false;- getCompanyProfile()- .then((p) => {- if (cancelled) return;- setCompanyName(p.name?.trim() || null);- if (!edited.current) setQuestion(firstQuestion(p.strategic_priorities, solo));- })- .catch(() => undefined);- listPeople()- .then((ps) => {- if (!cancelled) setPeopleCount(ps.length);- })- .catch(() => undefined);- return () => {- cancelled = true;- };- }, [solo]);-- const knows = [- companyName,- !solo && peopleCount !== null && peopleCount > 1 ? `your ${peopleCount} people` : null,- "this year's priorities",- ].filter(Boolean) as string[];- const knowsLine = `It knows ${knows.length > 1 ? `${knows.slice(0, -1).join(", ")} and ${knows[knows.length - 1]}` : knows[0]}.`;-- function ask() {- const text = question.trim();- if (!text) return;- remember(STORAGE_KEYS.firstQuestion, text);- onDone();- }-- return (- <div className="max-w-2xl mx-auto px-4 sm:px-6 py-12 sm:py-16 w-full">- <div className="text-4xl leading-none text-emerald-400" aria-hidden>- ✓- </div>- <h1 className="mt-3 text-2xl font-bold tracking-tight text-fg">Your Executive is ready</h1>- <p className="text-sm text-fg-muted mt-1">{knowsLine} Everything stays editable in Settings.</p>-- <form- onSubmit={(e) => {- e.preventDefault();- ask();- }}- className="mt-6 rounded-2xl border border-accent/30 bg-accent/5 p-4"- >- <label htmlFor="first-question" className="text-xs font-semibold uppercase tracking-wider text-accent">- Start here- </label>- <textarea- id="first-question"- value={question}- onChange={(e) => {- edited.current = true;- setQuestion(e.target.value);- }}- onKeyDown={(e) => {- if (e.key === "Enter" && !e.shiftKey) {- e.preventDefault();- ask();- }- }}- rows={2}- className="mt-2 w-full resize-none rounded-xl border border-line-strong bg-surface-elevated px-3 py-2 text-[15px] text-fg focus:outline-none focus:ring-2 focus:ring-accent/40"- />- <p className="mt-1 text-xs text-fg-muted">- Built from what you just saved. Change it, or ask anything else.- </p>- <div className="mt-3 flex flex-wrap items-center gap-x-4 gap-y-2">- <button type="submit" disabled={!question.trim()} className={buttonClass("primary", "md")}>- Ask your first question- </button>- <button- type="button"- onClick={onDone}- className="text-sm font-semibold text-accent hover:underline cursor-pointer"- >- Open Home instead- </button>- </div>- </form>-- <h2 className="mt-8 text-xs font-semibold uppercase tracking-wider text-fg-subtle">It can already</h2>- <ul className="mt-2 rounded-2xl border border-line bg-surface-elevated px-4">- <ExecEmailRow />- <li className="flex items-baseline gap-3 border-b border-line py-3 last:border-0">- <span aria-hidden className="w-5 flex-shrink-0 text-center">- ☀- </span>- <p className="min-w-0 text-[15px] text-fg-muted leading-snug">- <span className="font-semibold text-fg">Brief you.</span> Your first brief is on Home now, and a new one- is written every morning.- </p>- </li>- <li className="flex items-baseline gap-3 border-b border-line py-3 last:border-0">- <span aria-hidden className="w-5 flex-shrink-0 text-center">- ✎- </span>- <p className="min-w-0 text-[15px] text-fg-muted leading-snug">- <span className="font-semibold text-fg">Draft replies as you.</span>{" "}- <Link href={FIRST_RUN_LINKS.ownMailbox} className="font-semibold text-accent hover:underline">- Let it read your own mailbox- </Link>- . {OWN_MAILBOX_SENDS}- </p>- </li>- </ul>- </div>- );-}diff --git a/packages/ui/src/components/onboard/OnboardDraftReview.tsx b/packages/ui/src/components/onboard/OnboardDraftReview.tsxindex b6b6eef8..cfa0dcc5 100644--- a/packages/ui/src/components/onboard/OnboardDraftReview.tsx+++ b/packages/ui/src/components/onboard/OnboardDraftReview.tsx@@ -2,8 +2,9 @@import { useCallback, useEffect, useState, type ReactNode } from "react";import { useSession } from "next-auth/react";-import { ProfileSections } from "@/components/company-profile/ProfileSections";+import { ProfileSections, type SectionId } from "@/components/company-profile/ProfileSections";import OnboardDepartmentsDraft from "@/components/onboard/OnboardDepartmentsDraft";+import OnboardGoalsDraft from "@/components/onboard/OnboardGoalsDraft";import OnboardPeopleDraft from "@/components/onboard/OnboardPeopleDraft";import { profileWording } from "@/components/shell/navConfig";import { useWorkspace } from "@/components/workspace/WorkspaceContext";@@ -11,11 +12,37 @@ import {commitOnboardDraft,listDepartments,listPeople,+ updateWorkspace,type CompanyProfile,type OnboardDepartmentDraft,+ type OnboardGoalDraft,type OnboardPersonDraft,type OnboardTurn,+ type WorkspaceMode,} from "@/lib/api";+import { afterSaveSteps, checkNotes, goalsToSave, modeLine } from "@/lib/onboardSetup";++// Profile sections the review leaves out while they are empty: a long list of+// "Not set" reads like a form to fill in. They stay on Company Profile.+const OPTIONAL_SECTIONS: { id: SectionId; empty: (p: CompanyProfile) => boolean }[] = [+ { id: "mission", empty: (p) => !p.mission.trim() && !p.vision.trim() },+ {+ id: "customer",+ empty: (p) => !p.target_customer.profile.trim() && p.target_customer.pain_points.length === 0,+ },+ {+ id: "competitive",+ empty: (p) =>+ p.competitive_landscape.primary_competitors.length === 0 &&+ p.competitive_landscape.competitive_advantages.length === 0,+ },+ { id: "dependencies", empty: (p) => (p.vendors ?? []).length === 0 && (p.tickers ?? []).length === 0 },+ { id: "culture", empty: (p) => p.culture.values.length === 0 && p.culture.operating_principles.length === 0 },+ {+ id: "financials",+ empty: (p) => p.financials.burn_rate_monthly == null && p.financials.runway_months == null,+ },+];interface Props {turn: OnboardTurn;@@ -38,10 +65,22 @@ export default function OnboardDraftReview({const [departments, setDepartments] = useState<OnboardDepartmentDraft[]>(turn.draft_departments);+ // A goal the draft gave to the principal by name is theirs: blank, which+ // the owner picker shows as "you" and the server saves as the principal.+ const [goals, setGoals] = useState<OnboardGoalDraft[]>(() => {+ const me = turn.draft_people.find((p) => p.is_principal)?.full_name.trim().toLowerCase();+ return (turn.draft_goals ?? []).map((g) =>+ me && g.owner_name.trim().toLowerCase() === me ? { ...g, owner_name: "" } : g+ );+ });+ // Personal or Team was chosen from the text, never asked; the line at the+ // top shows it, and Change flips it here. It is saved with the rest.// Solo (one person, just for themselves): only "you" is shown and saved —// the drafted principal, else the first person drafted, else a blank row —// and no departments; the existing areas stay as they are.- const { mode, role: workspaceRole } = useWorkspace();+ const { mode: savedMode, role: workspaceRole, refresh } = useWorkspace();+ const [pickedMode, setPickedMode] = useState<WorkspaceMode | null>(null);+ const mode = pickedMode ?? savedMode;const solo = mode === "solo";// The profile's headings follow the role, as on /company-profile.const wording = profileWording(mode, workspaceRole.role_kind);@@ -92,6 +131,19 @@ export default function OnboardDraftReview({: people.filter((p) => p.full_name.trim());const principals = namedPeople.filter((p) => p.is_principal).length;const namedDepartments = solo ? [] : departments.filter((d) => d.title.trim());+ const goalsOut = goalsToSave(+ goals,+ namedPeople.map((p) => p.full_name),+ namedDepartments.map((d) => d.title)+ );+ const notes = checkNotes(turn.confidence_notes);+ const omit: SectionId[] = [+ // org_structure and this year's priorities are derived on save from the+ // people, departments and goals below; two editors would diverge.+ "org",+ "priorities",+ ...OPTIONAL_SECTIONS.filter((s) => s.empty(turn.draft!)).map((s) => s.id),+ ];const duplicateNames =new Set(namedPeople.map((p) => p.full_name.trim().toLowerCase())).size !==namedPeople.length;@@ -126,7 +178,18 @@ export default function OnboardDraftReview({setSaving(true);setError(null);try {- await commitOnboardDraft(turn.session_id, profile, namedPeople, namedDepartments, ownerEmail);+ if (mode !== savedMode) {+ await updateWorkspace({ mode });+ await refresh();+ }+ await commitOnboardDraft(+ turn.session_id,+ profile,+ namedPeople,+ namedDepartments,+ goalsOut,+ ownerEmail+ );onSaved();} catch (err) {setError((err as Error).message);@@ -137,25 +200,51 @@ export default function OnboardDraftReview({return (<div className="flex flex-col gap-4"><div>- <h1 className="text-lg font-semibold text-fg">Here’s what I understood</h1>+ <h1 className="text-lg font-semibold text-fg">+ {profile.name.trim() ? <>Here’s {profile.name.trim()}</> : <>Here’s what I understood</>}+ </h1><p className="text-sm text-fg-muted mt-0.5">- Nothing is saved yet. Edit anything that’s off, then save.+ This is what your Executive will know. Nothing is saved yet: edit anything that’s+ off, then save.</p></div>+ <div className="flex flex-wrap items-center gap-x-2 gap-y-1 rounded-xl border border-line bg-surface-elevated px-5 py-3 text-sm">+ <span className="text-fg">+ {modeLine(+ mode,+ people.map((p) => p.full_name)+ )}+ </span>+ <span className="text-fg-subtle" aria-hidden>+ ·+ </span>+ <button+ type="button"+ onClick={() => {+ const next: WorkspaceMode = solo ? "team" : "solo";+ // Team needs the roster: start it from "you" when the draft had only them.+ if (next === "team" && people.every((p) => !p.full_name.trim())) setPeople([me]);+ setPickedMode(next);+ }}+ disabled={saving}+ className="text-indigo-400 hover:text-indigo-300 transition-colors"+ >+ {solo ? "Change to a team" : "Change to just me"}+ </button>+ </div>+{turn.summary && (<div className="bg-surface-elevated border border-line rounded-xl px-5 py-4"><p className="text-sm text-fg whitespace-pre-wrap">{turn.summary}</p></div>)}- {turn.confidence_notes.length > 0 && (+ {notes.length > 0 && (<div className="bg-amber-500/10 border border-amber-500/20 rounded-xl px-5 py-4">- <p className="text-xs font-medium text-amber-400 uppercase tracking-wide mb-2">- I couldn’t determine these- </p>- <ul className="flex flex-col gap-1">- {turn.confidence_notes.map((note, i) => (+ <p className="text-sm font-semibold text-fg mb-2">Please check</p>+ <ul className="flex flex-col gap-1 list-disc pl-5">+ {notes.map((note, i) => (<li key={i} className="text-sm text-fg">{note}</li>@@ -166,16 +255,21 @@ export default function OnboardDraftReview({{answers}- {/* "org" is omitted: org_structure is derived from the two tables below- when you save, so editing it here would be a second source of truth. */}<ProfileSectionsprofile={profile}saving={false}onSave={mergeIntoDraft}- omit={["org"]}+ omit={omit}wording={wording}/>+ <OnboardGoalsDraft+ goals={goals}+ people={solo ? [] : people}+ departments={namedDepartments.map((d) => d.title.trim())}+ onChange={setGoals}+ />+{solo ? (<div className="bg-surface-elevated border border-line rounded-xl p-5"><h2 className="text-sm font-semibold text-fg">You</h2>@@ -224,6 +318,16 @@ export default function OnboardDraftReview({return valid.has(head) ? d : { ...d, head_person_name: "" };}));+ // Goal owners follow the same way; one whose owner is gone is yours.+ setGoals((gs) =>+ gs.map((g) => {+ const owner = g.owner_name.trim();+ if (!owner) return g;+ const moved = renamed.get(owner);+ if (moved) return { ...g, owner_name: moved };+ return valid.has(owner) ? g : { ...g, owner_name: "" };+ })+ );setPeople(next);}}/>@@ -255,7 +359,10 @@ export default function OnboardDraftReview({className="w-full rounded-lg border border-line-strong bg-surface-overlay px-3 py-2 text-sm text-fg placeholder-fg-subtle focus:outline-none focus:ring-2 focus:ring-indigo-500/50 transition-colors"/></div>- {!solo && (+ {/* Only when setup drafted departments (the description listed people+ by department). Otherwise there is nothing to review: departments+ are added later from the Goals page. */}+ {!solo && turn.draft_departments.length > 0 && (<OnboardDepartmentsDraftdepartments={departments}people={people}@@ -264,6 +371,21 @@ export default function OnboardDraftReview({/>)}+ <div className="bg-surface-elevated border border-line rounded-xl px-5 py-4">+ <p className="text-sm font-semibold text-fg mb-2">After you save, I’ll</p>+ <ul className="flex flex-col gap-1 list-disc pl-5">+ {afterSaveSteps(profile, goalsOut).map((step) => (+ <li key={step} className="text-sm text-fg">+ {step}+ </li>+ ))}+ </ul>+ <p className="text-xs text-fg-muted mt-3">+ Home then has a short list of things worth connecting. Everything here stays editable+ on the Company Profile, Goals and People pages.+ </p>+ </div>+{error && <p className="text-sm text-red-400">{error}</p>}{blocker && <p className="text-xs text-fg-muted">{blocker}</p>}@@ -273,14 +395,14 @@ export default function OnboardDraftReview({disabled={saving || blocker !== null}className="px-4 py-2 bg-indigo-500 hover:bg-indigo-600 disabled:opacity-40 text-white text-sm font-medium rounded-lg transition-colors">- {saving ? "Saving…" : "Save & finish setup"}+ {saving ? "Saving…" : "Save and start"}</button><buttononClick={onBackToConversation}disabled={saving}className="text-xs text-fg-muted hover:text-fg disabled:opacity-40 transition-colors">- Not quite — ask me more questions+ Back to questions</button></div></div>diff --git a/packages/ui/src/components/onboard/OnboardGoalsDraft.tsx b/packages/ui/src/components/onboard/OnboardGoalsDraft.tsxnew file mode 100644index 00000000..7b547c68--- /dev/null+++ b/packages/ui/src/components/onboard/OnboardGoalsDraft.tsx@@ -0,0 +1,109 @@+"use client";++import { type OnboardGoalDraft, type OnboardPersonDraft } from "@/lib/api";++const FIELD =+ "rounded-lg border border-line-strong bg-surface-overlay px-3 py-2 text-sm text-fg placeholder-fg-subtle focus:outline-none focus:ring-2 focus:ring-indigo-500/50 transition-colors";++interface Props {+ goals: OnboardGoalDraft[];+ /** Who can own a goal. Empty in Personal: every goal is yours. */+ people: OnboardPersonDraft[];+ /** Drafted department titles. Empty: no department picker. */+ departments: string[];+ onChange: (goals: OnboardGoalDraft[]) => void;+}++// "This year" on the review screen: each goal with its owner (and its+// department, only when setup drafted departments). Saved as goals, which+// are checked every week, and as this year's priorities on the profile.+export default function OnboardGoalsDraft({ goals, people, departments, onChange }: Props) {+ const named = people.filter((p) => p.full_name.trim());++ function update(i: number, patch: Partial<OnboardGoalDraft>) {+ onChange(goals.map((g, j) => (j === i ? { ...g, ...patch } : g)));+ }++ return (+ <div className="bg-surface-elevated border border-line rounded-xl p-5">+ <div className="flex items-baseline justify-between mb-1">+ <h2 className="text-sm font-semibold text-fg">This year</h2>+ <button+ type="button"+ onClick={() => onChange([...goals, { title: "", target: "", owner_name: "", department_title: "" }])}+ className="text-xs text-indigo-400 hover:text-indigo-300 transition-colors"+ >+ Add a goal+ </button>+ </div>+ <p className="text-xs text-fg-muted mb-4">+ Your goals. Each one is checked weekly, and any at risk shows in your brief.+ </p>++ {goals.length === 0 && (+ <p className="text-sm text-fg-subtle italic">No goals yet. Add one, or add them later on the Goals page.</p>+ )}++ <div className="flex flex-col gap-3">+ {goals.map((g, i) => (+ <div key={i} className="flex flex-col gap-2 pb-3 border-b border-line last:border-0 last:pb-0">+ <div className="flex items-center gap-2">+ <input+ value={g.title}+ onChange={(e) => update(i, { title: e.target.value })}+ placeholder="e.g. Launch the hosted version"+ aria-label="Goal"+ className={`flex-1 min-w-0 ${FIELD}`}+ />+ <button+ type="button"+ onClick={() => onChange(goals.filter((_, j) => j !== i))}+ aria-label={`Remove ${g.title || "goal"}`}+ className="text-xs text-fg-subtle hover:text-red-400 px-1 transition-colors"+ >+ ✕+ </button>+ </div>+ {(named.length > 1 || departments.length > 0) && (+ <div className="flex flex-col sm:flex-row gap-2">+ {named.length > 1 && (+ <select+ value={g.owner_name}+ onChange={(e) => update(i, { owner_name: e.target.value })}+ aria-label="Owner"+ className={`flex-1 ${FIELD}`}+ >+ {/* Blank means the person marked "This is me". */}+ <option value="">Owner: you</option>+ {named+ .filter((p) => !p.is_principal)+ .map((p) => (+ <option key={p.full_name} value={p.full_name}>+ Owner: {p.full_name}+ </option>+ ))}+ </select>+ )}+ {departments.length > 0 && (+ <select+ value={g.department_title}+ onChange={(e) => update(i, { department_title: e.target.value })}+ aria-label="Department"+ className={`flex-1 ${FIELD}`}+ >+ <option value="">No department</option>+ {departments.map((d) => (+ <option key={d} value={d}>+ {d}+ </option>+ ))}+ </select>+ )}+ </div>+ )}+ </div>+ ))}+ </div>+ </div>+ );+}diff --git a/packages/ui/src/lib/api.ts b/packages/ui/src/lib/api.tsindex 5ea7366b..1807b35a 100644--- a/packages/ui/src/lib/api.ts+++ b/packages/ui/src/lib/api.ts@@ -412,6 +412,15 @@ export interface OnboardDepartmentDraft {authority_level: string;}+/** One of this year's goals. owner_name is a drafted person's name (blank =+ * the owner of the workspace); department_title names a drafted department. */+export interface OnboardGoalDraft {+ title: string;+ target: string;+ owner_name: string;+ department_title: string;+}+export interface OnboardTurn {session_id: string;phase: "question" | "draft";@@ -419,9 +428,12 @@ export interface OnboardTurn {max_questions: number;question: string | null;question_hint: string | null;+ /** Suggested answers, shown as buttons to tap. */+ question_options: string[];draft: CompanyProfile | null;draft_people: OnboardPersonDraft[];draft_departments: OnboardDepartmentDraft[];+ draft_goals: OnboardGoalDraft[];confidence_notes: string[];summary: string | null;}@@ -459,7 +471,7 @@ export async function startOnboardInterview(}// What the first free-text description already says. Every field is null when-// the text does not say; the confirm step asks for those.+// the text does not say; the interview asks for what matters.export interface OnboardUnderstanding {mode: WorkspaceMode | null;role_kind: RoleKind | null;@@ -539,6 +551,7 @@ export async function commitOnboardDraft(profile: CompanyProfile,people: OnboardPersonDraft[],departments: OnboardDepartmentDraft[],+ goals: OnboardGoalDraft[],/** The principal's sign-in email, confirmed on the review screen. Blank = don't set one. */ownerEmail = ""): Promise<CompanyProfile> {@@ -550,6 +563,7 @@ export async function commitOnboardDraft(profile,people,departments,+ goals,owner_email: ownerEmail.trim() || null,}),});diff --git a/packages/ui/src/lib/onboardSetup.ts b/packages/ui/src/lib/onboardSetup.tsnew file mode 100644index 00000000..ea808bf3--- /dev/null+++ b/packages/ui/src/lib/onboardSetup.ts@@ -0,0 +1,92 @@+// Pure helpers for the setup flow (/onboard), kept out of the components so+// scripts/onboardSetup.test.mjs can check them under+// `node --experimental-strip-types` (type-only imports).++import type {+ CompanyProfile,+ OnboardGoalDraft,+ OnboardUnderstanding,+ WorkspaceMode,+ WorkspaceUpdate,+} from "@/lib/api";++/** The mode setup saves without asking: what the text says, else Personal+ * (nothing is lost, and adding a teammate later switches to Team). */+export function inferredMode(u: OnboardUnderstanding | null): WorkspaceMode {+ return u?.mode ?? "solo";+}++/** The role fields the text gave, for a Personal workspace. Only what was+ * said: a field the text left out keeps whatever is already saved. */+export function understoodRole(u: OnboardUnderstanding | null): WorkspaceUpdate {+ const out: WorkspaceUpdate = {};+ if (u?.role_kind) out.role_kind = u.role_kind;+ if (u?.role_title) out.role_title = u.role_title;+ if (u?.reports_to) out.reports_to = u.reports_to;+ return out;+}++const SHOWN_NAMES = 4;++/** The one line the review shows in place of a "Who is this for?" question. */+export function modeLine(mode: WorkspaceMode, names: string[]): string {+ if (mode === "solo") return "Set up just for you";+ const named = names.map((n) => n.trim()).filter(Boolean);+ if (named.length === 0) return "Set up for your team";+ const shown = named.slice(0, SHOWN_NAMES);+ const rest = named.length - shown.length;+ return `Set up for your team: ${shown.join(", ")}${rest > 0 ? ` and ${rest} more` : ""}`;+}++// A snake_case word ("annual_revenue_arr") is a field name, not a sentence.+const FIELD_NAME = /\b[a-z]+(?:_[a-z0-9]+)+\b/;++/** The "Please check" lines: the drafter's notes, minus any that still+ * speak in field names (the prompt asks for plain sentences; this is the+ * backstop) and blanks. */+export function checkNotes(notes: string[]): string[] {+ return notes.map((n) => n.trim()).filter((n) => n && !FIELD_NAME.test(n));+}++/** What happens after Save, as the review lists it under "After you save,+ * I'll". Research always runs once setup is saved+ * (api/routes/onboarding.py, _fire_post_onboarding_research). */+export function afterSaveSteps(profile: CompanyProfile, goals: OnboardGoalDraft[]): string[] {+ const steps: string[] = [];+ const competitors = profile.competitive_landscape?.primary_competitors ?? [];+ steps.push(+ competitors.length > 0+ ? "Research your market and competitors, and put what's worth knowing in your brief"+ : "Research your competitors and suggest a shortlist in your brief",+ );+ if (goals.some((g) => g.title.trim())) {+ steps.push("Check your goals every week and flag any that are at risk in your brief");+ }+ const watched = [...(profile.vendors ?? []), ...(profile.tickers ?? [])].filter((v) => v.trim());+ if (watched.length > 0) {+ steps.push(`Watch the status pages and news for ${watched.join(", ")}`);+ }+ return steps;+}++/** Goals as the commit sends them: named ones only, trimmed, and any owner+ * or department that no longer exists on the review screen cleared (the+ * server would refuse the whole save over it). */+export function goalsToSave(+ goals: OnboardGoalDraft[],+ people: string[],+ departments: string[],+): OnboardGoalDraft[] {+ const names = new Set(people.map((p) => p.trim().toLowerCase()).filter(Boolean));+ const titles = new Set(departments.map((d) => d.trim().toLowerCase()).filter(Boolean));+ return goals+ .filter((g) => g.title.trim())+ .map((g) => ({+ title: g.title.trim(),+ target: g.target.trim(),+ owner_name: names.has(g.owner_name.trim().toLowerCase()) ? g.owner_name.trim() : "",+ department_title: titles.has(g.department_title.trim().toLowerCase())+ ? g.department_title.trim()+ : "",+ }));+}