Annsa

Added value: +"Where to look: feedback (default), annsa (the docs), surveys, customer, accounts, templates or transcript."

Previous value: -"Narrows `customer` to one company"New value: +"With customer: which company, if the name matches more than one."

Previous value: -"With about='customer': a name, email or id"New value: +"With about=customer: their name, email or id."

Previous value: -"The text you are about to file; returns the nearest priorities"New value: +"The text you're about to file. Returns the nearest priorities, each judged same, possibly or different."

Previous value: -"With do_we_have: a customer's words or an agent's finding"New value: +"With do_we_have: customer (a customer's words) or agent (an agent's finding)."

Previous value: -"The version of your saved copy of the file"New value: +"With skill: the version of the copy you already have, so you only get it if it changed."

Added value: +"For a long transcript: which page, starting at 1."

Added value: +"What you want to know, in a sentence."

Previous value: -"Returns a skill file instead of an answer; no question needed"New value: +"Returns a skill file instead of an answer: working, spec-skill, decisions or welcome. No question needed."

Previous value: -"t12 or its id; long ones come in pages"New value: +"With about=transcript: which call, like t12 or its id. Long ones come in pages."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"set_stage: Sets the planning stage. correct: Fixes a title, detail or ruled-out entry, keeping the old version. retract: Unfiles feedback and marks a ship not shipped. regroup: Moves priorities between groups; the reply carries the undo. merge: Folds duplicates into `into`; their numbers are retired. Refuses shipped priorities. bet: Manages bets and places priorities on them. cut: Closes a release and writes its changelog. pick: Sets a person's short list. approve_findings: Lets through agent findings held back by the daily limit; the owner's. view: Saves a board view; an agent's is a suggestion."New value: +"set_stage: Move a priority to open, ready or closed. assign: Set who a spec is assigned to. A person's action. correct: Fix a title, detail or ruled-out entry. The old wording is kept. park: Take a priority off the list, with a reason, until someone unparks it. A person's action. unpark: Put a parked priority back on the list. retract: Unfile feedback, or mark a shipped priority as not shipped. A person's action; without confirm it only shows the plan. regroup: Move priorities between groups; the reply carries the undo. merge: Fold duplicate priorities into one. Their numbers are retired for good. Shipped priorities can't be merged. bet: List, create, update, delete or reorder bets, or place priorities on one. op says which. direction: Read, list, update or comment on a Direction page. op says which. set_severity: Set a priority's severity, P0 to P3. From an agent it's a proposal and needs a reason; a person's typed severity wins. cut: Close a release and write its changelog. op=publish marks its rows Shared. pick: Set a person's short list of priorities, in order. approve_findings: Let through agent findings held back by the daily limit. The owner's action. view: Save a board view. An agent's is a suggestion until a person keeps or dismisses it."

Previous value: -"How much of your effort this bet should get, 0 to 100."New value: +"For a bet: how much of your effort it should get, 0 to 100."

Previous value: -"Retract is refused once customers were told, unless true"New value: +"For retract: true to retract even after customers were told."

Previous value: -"An agent client or a member's email; empty unassigns"New value: +"Who gets it: an agent (like claude-code or codex) or a person's email. Empty unassigns."

Added value: +"For cut: the release date."

Added value: +"For bet update, delete or place: which bet."

Previous value: -"The group this bet counts toward, like Product or GTM."New value: +"For a bet: the group it counts toward, like Product or GTM."

Previous value: -"For cut op=publish."New value: +"For cut op=publish: changelog or marketing."

Added value: +"For direction op=comment: your words."

Previous value: -"Without it retract only shows the plan"New value: +"For retract: true to do it. Without it the reply only shows what would change."

Previous value: -"What it should say now."New value: +"For correct: what it should say now."

Previous value: -"'category' is a person's; the old wording is kept"New value: +"For correct: 'title', 'detail', 'ruled_out' or 'category'. 'category' is a person's; the old wording is kept."

Added value: +"For direction approve or dismiss: which page."

Previous value: -"At least 50 characters"New value: +"For direction update: the page text, at least 50 characters."

Added value: +"For merge: the priorities being folded into `into`."

Added value: +"For direction op=update: the evidence behind the change."

Previous value: -"Only change it if the stage is still this."New value: +"For set_stage: only change it if the stage is still this: open, ready or closed."

Added value: +"For retract: the feedback rows to unfile."

Added value: +"For merge: the priority that stays."

Previous value: -"For regroup: the new area for each priority."New value: +"For regroup: the new group for each priority."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Added value: +"For a new bet: what the measure is called."

Previous value: -"For a written bet: the number now (a person's)."New value: +"For a written bet: the number now. A person's."

Previous value: -"For action 'bet': 'list', 'create' (title + metric_label; last for agents), 'update' (bet_id), 'delete', 'reorder', 'place' (priority_ids). For action 'direction': 'get', 'list', 'update' (document_text + reason), 'comment' (thread_id replies), 'approve' and 'dismiss' (document_id + reason; a person's). For action 'cut': publish takes release_number, channel and the live annsa.ai url; it marks the release's rows Shared and sends nothing. For action 'view': 'keep' and 'dismiss' are a person's."New value: +"For bet: 'list', 'create' (title and metric_label; agents last), 'update' (bet_id), 'delete', 'reorder' (bet_ids) or 'place' (priority_ids). For direction: 'get', 'list', 'update' (document_text and reason), 'comment' (thread_id), or 'approve' and 'dismiss' (document_id and reason; a person's). For cut: 'publish', with release_number, channel and the live annsa.ai url. It marks the release's rows Shared and sends nothing. For view: 'list', 'create', 'update', 'delete', or 'keep' and 'dismiss' (a person's)."

Previous value: -"For a category fix, the old category; for a ruled-out fix, the entry being replaced."New value: +"For correct: the old category, or the ruled-out entry being replaced."

Previous value: -"'none' clears the owner"New value: +"For bet update: the owner's user id; none clears it."

Added value: +"The priority this one sits under. For bet op=place: the bet."

Previous value: -"Why it's parked (required, up to 500 characters)."New value: +"For park: why, up to 500 characters."

Previous value: -"Email or user id of the member whose words they are"New value: +"The email or user id of the person whose words they are."

Previous value: -"The person's exact words, not a summary; recorded as their quote"New value: +"The person's exact words, not a summary. Recorded as their quote."

Added value: +"For bet op=place or direction: the priorities."

Added value: +"For view: the question the saved view answers."

Added value: +"For direction op=comment: the line you're replying to."

Previous value: -"Required for takeover, release, retract, correct and an agent's severity"New value: +"Why, in a line. Needed for retract, correct, merge and an agent's severity."

Previous value: -"annsa.priorities args"New value: +"For view: the annsa.priorities arguments the view runs."

Added value: +"For cut op=publish: which release."

Previous value: -"0 broken, 1 poor, 2 could be better, 3 minor. From an agent: a proposal until a person types the P."New value: +"0 broken, 1 poor, 2 could be better, 3 minor. From an agent it's a proposal until a person sets it."

Added value: +"For cut: the start of the window the release covers."

Previous value: -"The spec id from claim or annsa.spec (build_…)."New value: +"For assign and retract: the spec id."

Added value: +"For set_stage: open, ready or closed."

Added value: +"For direction op=comment: the thread to reply in."

Added value: +"For bet or view: its title."

Added value: +"For cut: the end of that window."

Added value: +"For cut op=publish: the live changelog link."

Previous value: -"The saved view to update or delete, from view op=list."New value: +"For view update or delete: which view, from view op=list."

Previous value: -"Why this bet matters now, in a line."New value: +"For a bet: why it matters now, in a line."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"approve: Writes another writer's proposed section text. dismiss: Keeps the current section text. answer: Records a person's answer, typed or relayed with person_words; an option that names an act runs it. direct: A person steers the agent holding a priority; it reads the words on its next reply, once. choose: Records a person's pick on a Needs you question: one option, or neither. sign: Records a person's decision. principle: Decision memory; propose applies at once."New value: +"approve: Accept another writer's proposed section text, by update_id. dismiss: Turn down a proposed section text; the current text stays. answer: Record a person's answer to the open ask on a priority. An agent passes on the person's exact words with person_words. Picking an option labelled Merge, Ship it, Park it or Retract the ship runs it; free text is only recorded. ask_back: A person asks the agent a question back on a priority. A person's action. direct: A person gives the agent holding a priority a direction; the agent reads it on its next reply. A person's action. choose: Record a person's pick on a Needs you question (send disagreement_id). One option, or neither. A person's action. sign: Record a person's decision on a priority (do_this, do_nothing, park or dismiss), with because. A person's action. principle: Your decision memory. Agents can list principles or propose one; a person writes, retires or reactivates them. A proposal applies at once."

Added value: +"For choose: the option picked, or neither. For principle: the answer behind the rule."

Added value: +"The person who decided, when you're passing on their words."

Added value: +"For scope=area: the part of the product the rule covers."

Added value: +"For sign: why, in a sentence."

Added value: +"For sign: do_this, do_nothing, park or dismiss."

Previous value: -"YYYY-MM-DD"New value: +"For sign: when this decision expires, like 2026-12-31."

Previous value: -"The id of the Needs you question being answered, from the Needs you list."New value: +"For choose: the id of the Needs you question, from the Needs you list."

Previous value: -"YYYY-MM-DD: when to check wrong_if"New value: +"For sign: when to check wrong_if, like 2026-12-31."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Previous value: -"Needs answered_by"New value: +"For sign: what the agent may decide on its own from here. Needs answered_by."

Added value: +"For sign: what the agent has to bring back to a person."

Previous value: -"For action 'principle': 'propose' (an agent's), 'write', 'retire', 'activate' (a person's), 'list'."New value: +"For principle: 'list', 'propose' (an agent's), or 'write', 'retire' and 'activate' (a person's)."

Added value: +"For answer: the option's label."

Previous value: -"Email or user id of the member whose words they are"New value: +"The email or user id of the person whose words they are."

Previous value: -"The person's exact words, not a summary; recorded as their quote"New value: +"The person's exact words, not a summary. Recorded as their quote."

Previous value: -"The rule that would settle this class next time"New value: +"For principle: the rule, in a sentence."

Previous value: -"The principle to retire or activate, from principle op=list."New value: +"For principle retire or activate: which one, from principle op=list."

Previous value: -"When to look at this decision again, in a sentence."New value: +"For sign: when to look at this decision again, in a sentence."

Previous value: -"One 'label — why' per entry"New value: +"Options you ruled out, one per entry as 'option — why'."

Previous value: -"area needs `area`; area or company also writes a principle"New value: +"How far the decision reaches: priority, area (needs area) or company. area or company also writes a principle."

Previous value: -"The note or principle this replaces."New value: +"For principle: the principle this replaces."

Previous value: -"For do_this: the option you chose, in a few words (up to 200 characters)."New value: +"For sign with do_this: the option you chose, in a few words."

Added value: +"For direct or ask_back: the words."

Added value: +"For principle: the rule's title."

Added value: +"For principle: what the rule is about."

Previous value: -"The proposed change to approve or dismiss, from annsa.spec's updates."New value: +"For approve or dismiss: the proposed change, from annsa.spec's updates."

Previous value: -"What the agent may keep doing until the answer comes."New value: +"For sign: what the agent may keep doing until the answer comes."

Added value: +"For sign: what would show the decision was wrong."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"submit: Adds one feedback row; an agent's is a finding. With priority_number it attaches there; with stand_alone it proposes a new priority. report: Tell the Annsa team about a bug, gap or idea in Annsa. Needs `text`; `title` names the tool and action. The Annsa team reads it, so keep your customers' data out. add_page: Makes a page from a template, private to you; returns its link. add_template: Saves a template (title, text, when_to_use), private to you; {blanks} are its fields. connect_goal: Answer the welcome's optional question: what the person wants from Annsa, up to 300 characters. Goes to the Annsa team, not your board. import_direction: Two steps: preview a strategy doc's lines, then add the ones you pick as ideas. add_priority: Adds a priority from a title and detail, with no customer words."New value: +"submit: File what a customer said or what you found. With priority_number it attaches to that priority; with stand_alone it starts a new one. report: Tell the Annsa team about a bug, gap or idea in Annsa itself. Send text; title names the tool and action. Keep your customers' data out of it. add_page: Make a page from a template, private to you. Returns the link. add_template: Save a template with a title, text and when to use it. {blanks} in the text become its fields. connect_goal: Answer the welcome's optional question about what you want from Annsa, up to 300 characters. Goes to the Annsa team, not your board. add_transcript: Add a call transcript as customer feedback. What the customer asked for is filed from it. A person's action. import_direction: Import a strategy doc in two steps. First preview its lines, then add the ones you pick as ideas. A person's action. add_priority: Add a priority from a title and detail, with no customer words."

Previous value: -"Raw scanner output is refused unless someone checked it"New value: +"True if a person checked this scanner finding; raw scanner output is refused."

Previous value: -"For add_transcript (default other)."New value: +"For add_transcript: interview, sales_call, support or other (the default)."

Added value: +"Their email, if you have it."

Added value: +"The customer who said it."

Previous value: -"The longer words of the row, up to 20000 characters"New value: +"For add_priority: the longer words of the row, up to 20000 characters."

Previous value: -"The Direction page to save into; leave it out for a new page."New value: +"Which Direction page. Leave it out for a new page, or the Overview."

Previous value: -"The page's title (kept if you send none)."New value: +"The page's title. Kept if you send none."

Previous value: -"At least 50 characters"New value: +"For import_direction: the document, at least 50 characters."

Previous value: -"The words, verbatim"New value: +"For submit: what was said, word for word."

Added value: +"For add_page: a value for each {blank} in the template."

Previous value: -"Who said it"New value: +"Who said it: customer, team, agent, idea or roadmap."

Previous value: -"Files it without the check for a matching priority."New value: +"For submit and add_priority: file it even if a matching priority exists."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Added value: +"For submit from a scanner: the line number."

Previous value: -"Where it came from, like a ticket or file:line; the same one again adds to the same row."New value: +"Where it came from, like a ticket or file:line. The same one again adds to the same row."

Added value: +"The priority this one sits under. For bet op=place: the bet."

Added value: +"For submit from a scanner: the file."

Previous value: -"For import_direction: the id phase 1 returned, to land phase 2"New value: +"For import_direction: the id the preview returned, to add the lines you picked."

Previous value: -"Required for takeover, release, retract, correct and an agent's severity"New value: +"For add_priority: why you're filing it, in a line."

Previous value: -"A followable link (PR, URL, log line); findings need one"New value: +"A link someone can follow: a PR, URL or log line. A finding needs one."

Previous value: -"An amount ('499') or a tier word"New value: +"What the customer pays: an amount like 499, or a tier word."

Previous value: -"When ___, ___ can't ___ because ___; sent on the submit. Max 300 characters"New value: +"The cause in one line, shaped 'When ___, ___ can't ___ because ___'. Up to 300 characters."

Added value: +"For submit from a scanner: the rule that fired."

Previous value: -"Titles to land in phase 2; omitted lands all"New value: +"For import_direction: which lines to add, by title. Leave it out to add all."

Previous value: -"0 broken, 1 poor, 2 could be better, 3 minor. From an agent: a proposal until a person types the P."New value: +"0 broken, 1 poor, 2 could be better, 3 minor. From an agent it's a proposal until a person sets it."

Previous value: -"Why this P; sent on the submit with severity"New value: +"Why this severity, in a line."

Previous value: -"Where the feedback came from, like intercom."New value: +"Where the feedback came from, like intercom. Defaults to mcp."

Previous value: -"Makes a new priority from this row; not with priority_number."New value: +"For submit: start a new priority from this row. Not with priority_number."

Added value: +"For add_page: the template's key, from annsa.ask about=templates."

Added value: +"For report, add_template, connect_goal and add_transcript: the words. submit takes feedback."

Previous value: -"One of the workspace's themes"New value: +"The group it belongs to, one of the workspace's themes."

Added value: +"For add_priority, add_page, add_template or report: the title."

Previous value: -"What kind of row it is."New value: +"Which kind: bug, feature, improvement, praise or question."

Previous value: -"On a shipped priority: still_broken or fixed. On release, false_positive does NOT close the row."New value: +"For submit on a shipped priority: still_broken or fixed. false_positive doesn't close the row."

Added value: +"For add_template: when an agent should pick this template."

Previous value: -"Also writes a ready spec and returns its spec_id."New value: +"For submit and add_priority: also write a ready spec and return its spec_id."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"A bet id, or 'unplaced' for the proposal inbox"New value: +"Only priorities on this bet, by id. unplaced lists the ones no bet has yet."

Previous value: -"Long lists, comma separated: agents, to_verify, in_build, unrated_filings or all."New value: +"Extra lists to include, comma separated: agents, to_verify, in_build, unrated_filings or all."

Added value: +"Include parked priorities too."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Previous value: -"Max 100"New value: +"How many to return. Defaults to 5, max 100."

Previous value: -"Only the priorities you hold; the reply carries `needs_me`."New value: +"Only the priorities you're working on. The reply also counts what needs you."

Added value: +"A person's email. With picked, whose picked list."

Previous value: -"A person's picked list; `person` says whose"New value: +"Only the priorities a person picked; person says whose."

Removed value: -"volume"

Previous value: -"retention is unhappy customers, delight is praise, feature is new-feature asks."New value: +"Leave it out for the board's order, the same as the app. Or rank by volume, urgency, revenue, retention (unhappy customers), delight (praise) or feature (new-feature asks)."

Added value: +"Only the priorities in this release."

Previous value: -"enterprise or smb"New value: +"Customer size: all (default), enterprise or smb."

Previous value: -"Default: the ranked active priorities (open, ready, building)."New value: +"open, ready, building, shipped, shared or closed. Leave it out for the ranked active priorities: open, ready and building."

Previous value: -"Matches a priority's wording and its feedback"New value: +"A word or phrase to search for in priorities and their feedback."

Added value: +"Only this kind: bug, feature, improvement, praise or question."

Previous value: -"radar, brief or a saved view"New value: +"A saved view by name, or radar (what's rising, shipped and quiet) or brief."

Added value: +"A summary of the week instead of the list: what's rising, what shipped, what went quiet."

Previous value: -"How far back the radar looks."New value: +"How far back the radar looks: 7d, 14d or 30d. Defaults to 7d."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"share: Emails or notifies the customers who asked that it shipped. An email cannot be unsent."New value: +"share: Email or notify the customers who asked that it shipped. An email cannot be unsent."

Previous value: -"Who to tell: widget is in the app; both is the default; skip tells nobody."New value: +"Who to tell: email, widget (in the app), both (the default) or skip (nobody)."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Previous value: -"Your own words, instead of the default note."New value: +"For share: your own words, instead of the default note."

Previous value: -"The spec id from claim or annsa.spec (build_…)."New value: +"The shipped spec, by id."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Added value: +"Export many specs at once, filtered by spec_ids, stage or since."

Previous value: -"With batch: list is one line each; full is the whole spec."New value: +"With batch: list (one line each, the default) or full (the whole spec)."

Previous value: -"repo_rules adds the repo's agent files."New value: +"One extra: repo_rules (the repo's agent files), decision (the Decision text), memory (related memory) or verification (the verify receipts)."

Previous value: -"Direction first screens"New value: +"Add the opening of each Direction page that applies."

Previous value: -"Adds Evidence and direction, Files to touch and the quotes"New value: +"Add the Evidence and direction and Files to touch sections, and the customer quotes."

Added value: +"Add related answers and what earlier work learned."

Previous value: -"The rules in full"New value: +"Add the full text of the principles that apply."

Added value: +"Add an analysis of the problem from the feedback and history."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Previous value: -"Max 50"New value: +"With batch: how many specs. Defaults to 10, max 50."

Added value: +{
+  "description": "With batch: only specs that reached staging, main or prod, newest first.",
+  "enum": [
+    "staging",
+    "main",
+    "prod"
+  ],
+  "type": "string"
+}

Previous value: -"ISO date or 'last month'"New value: +"A date like 2026-11-06, or 'last month'."

Previous value: -"With batch: only specs at this stage."New value: +"With batch: only specs at this stage: open, ready, building, shipped, shared or closed."

Added value: +"With batch: the spec's status: open, ready, building, shipped, shared, closed or done. Not the board stage; that's stage."

Added value: +"A date like 2026-11-06. With since, the window."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}

Previous value: -"ship: Records completion and deploys nothing; retract undoes it. claim: Claims the priority, starts a short spec and returns holder_key. note: Appends a note; a correction is a new note with supersedes. dispatch: Posts @claude on the priority's linked GitHub issue; the person stays assigned. unclaim: Puts the work down; nothing is marked shipped. close: Closes a priority in one call, claim or PR and all; it stays under Closed. write: Replaces one spec section as a new version; text for a section that is not yours is recorded as an update. verify: Records how a line was checked, by proof kind. No ship gate. escalate: Raises a three-line ask in a person's Needs you; decides nothing. Needs `checked`: a fact you can check is not asked. One open ask per priority; the same topic updates it. resolve: Answers a routine ask the way your principles already decide it; returns use_id for undo. undo: Puts back your own resolve by use_id."New value: +"ship: Mark the priority shipped. It only records it; nothing deploys. retract undoes it. claim: Take a priority to work on. Starts a spec if there isn't one and gives you spec_id and holder_key. note: Leave a note on the spec about what you did or found. To correct an earlier note, send a new one with supersedes. handoff: Pass the work to another agent or a person, with a note for them. dispatch: Post @claude on the priority's GitHub issue so Claude picks it up there. The person stays assigned. A person's action. unclaim: Unclaim a priority so another agent can pick it up. close: Close the priority in one call, even if it's claimed or has a PR. It stays under Closed with your reason. write: Replace one section of the spec with new text. If the section belongs to someone else, your text is saved as a proposed change for them. verify: Record how you checked one Done When line, by proof kind: a test id for (proof: test), a URL for staging or prod. It doesn't block ship. escalate: Ask a person to decide something; it lands in their Needs you. Send topic, ask, trigger, principle and checked. One open ask per priority; the same topic updates it. resolve: Settle a routine question, like 'is this a duplicate of #N?', the way your principles already decide it. Each outcome fits one kind of question. Returns use_id so you can undo. undo: Reverse your own resolve, by use_id."

Added value: +"For escalate: the question you want the person to answer, in one line."

Previous value: -"An agent client or a member's email; empty unassigns"New value: +"Who gets it: an agent (like claude-code or codex) or a person's email. Empty unassigns."

Previous value: -"What you checked before asking, as \"what — result\"."New value: +"For escalate: what you checked before asking, as 'what — result'."

Previous value: -"Which Done When line, counting from 1 as annsa.spec does."New value: +"For verify: which Done When line, counting from 1 as annsa.spec lists them."

Added value: +"For resolve: the rows the ask is about."

Added value: +"For verify: how you checked the line: test, staging, prod or manual."

Previous value: -"claim: adds related_answers, learned"New value: +"For claim: add related answers and what earlier work learned."

Previous value: -"claim: the rules in full"New value: +"For claim: add the full text of the principles that apply."

Added value: +"For resolve move_row: the priority the row moves to."

Added value: +"What to call this session, up to 60 characters. Shown as who did it."

Added value: +"For ship or escalate: what the next agent should know."

Added value: +{
+  "description": "For claim and ship: the model the agent runs on, like claude-opus-5-5. Shown beside its name.",
+  "type": "string"
+}

Previous value: -"What should happen next on this priority."New value: +"For note or handoff: what should happen next on this priority."

Previous value: -"Short labels, for a decision"New value: +"For escalate: the choices, as short labels."

Added value: +"For resolve: keep_both (a merge ask), close_settled, dismiss_came_back or move_row (a stray row). Each fits one kind of ask."

Previous value: -"GitHub PR URL; body has 'Annsa-Spec: <spec id>'"New value: +"The GitHub PR link. Its body carries 'Annsa-Spec: <spec id>'."

Previous value: -"The rule that would settle this class next time"New value: +"For escalate: the rule that would let an agent settle this kind of question next time."

Previous value: -"For escalate: decision picks between options, help needs a person's hand, gap is something missing, update is just telling."New value: +"For escalate: decision (pick between options), help (needs a person's hand), gap (something missing) or update (just telling)."

Previous value: -"On a note: how long you expect to be quiet, in minutes."New value: +"For note: how long you expect to be quiet, in minutes."

Previous value: -"Required for takeover, release, retract, correct and an agent's severity"New value: +"Why, in a line. Needed for takeover, close, unclaim, handoff, and a park or retract request."

Previous value: -"Needs reason; a person decides."New value: +"For escalate: park or retract, to ask the person for one. Needs reason."

Previous value: -"One 'label — why' per entry"New value: +"Options you ruled out, one per entry as 'option — why'."

Previous value: -"Proposes a merge for a person to answer"New value: +"For escalate: the priority this looks like a duplicate of; a person decides the merge."

Previous value: -"The spec section: the_problem, evidence_and_direction, what_to_build, files_to_touch or done_when. On a note: build, how_it_works or spec."New value: +"For write: the_problem, evidence_and_direction, what_to_build, files_to_touch or done_when. For note: build, how_it_works or spec."

Previous value: -"True for a security question"New value: +"For escalate: true if this is a security question."

Previous value: -"True when the ask is \"ship this PR\"; needs pr_url."New value: +"For escalate: true when the ask is 'ship this PR'. Needs pr_url."

Previous value: -"The spec id from claim or annsa.spec (build_…). Required for verify."New value: +"The spec id from claim, or build_id from annsa.spec. Required for verify."

Previous value: -"The note or principle this replaces."New value: +"For note: the note this replaces. For principle: the principle this replaces."

Previous value: -"Another session of your own client counts as someone else; needs a reason"New value: +"Claim or unclaim work someone else holds. Needs reason. Another session of your own counts as someone else."

Added value: +"For escalate: what the ask is about, in a few words."

Previous value: -"Anything else is settled, not raised."New value: +"For escalate: why this needs a person: scope, customers, principle or options. Anything else is settled, not raised."

Added value: +"For undo: the id resolve gave you."

Previous value: -"On a shipped priority: still_broken or fixed. On release, false_positive does NOT close the row."New value: +"For unclaim: false_positive marks the finding wrong. still_broken and fixed belong to annsa.file submit."

Previous value: -"A test id (path::test) or a URL that shows the check."New value: +"For verify: a test id (path::test) or a URL that shows the check."

Previous value: -nullNew value: +{
+  "properties": {
+    "message": {
+      "type": "string"
+    },
+    "record": {
+      "type": "object"
+    },
+    "success": {
+      "type": "boolean"
+    },
+    "who": {
+      "type": "object"
+    }
+  },
+  "type": "object"
+}