Propose a rule for the project’s memory
const url = 'https://beta-api.plune.ai/v1/memory/candidates';const options = { method: 'POST', headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'}, body: '{"source":{"kind":"answer","id":"6f1c2b3a-0000-4000-8000-0000000021a9"}}'};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request POST \ --url https://beta-api.plune.ai/v1/memory/candidates \ --header 'Authorization: Bearer <token>' \ --header 'Content-Type: application/json' \ --data '{ "source": { "kind": "answer", "id": "6f1c2b3a-0000-4000-8000-0000000021a9" } }'A candidate rule (#2198), on request: nothing proposes one by itself. From an answer to a question (answer: the question, answered), from the reason of a decline (decline: a proposal, a discovered test, an endpoint, a strategy, a report or a result a person rejected with a reason), from the action a person picked on a failed test (failure_choice: a question on a failed test answered with an action, which is no answer source: one record, one source), or from a lesson of a report (report_lesson: an approved report, named <its review item>:<the lesson’s questionId>). Known limit: a lesson is an answered question too, so the same answer can be proposed as answer and as report_lesson, two sources and two candidates. It becomes the eighth kind of queue entry, pending, the caller its author, logged as a create of the item; it is a rule only once a person approves it in a session (PATCH /v1/review-items/{id} with approved; an API token is 403 session_required). The source is its identity: proposed again, in any words, the answer is 200 with the entry already there, waiting or decided, and nothing is written. Without text the words are made from the record; an answer to an access question is never copied (400 text_required). A person or a write token. A waiting candidate takes a place under the review queue’s ceiling pendingReviewItems, as a question does: on a full queue a new source is 429 with the quota’s figures, and the one already there is still 200. Accepting it is asked against memoryRulesPerProject (the rules in force; a retired one is not counted): 200 on every plan, and 0 on Free (#2207, memory in the platform is on the paid plans; a 429 on the approval only once the operator sets TIER_LIMITS_ENFORCE=true). The words are masked as a failure’s text is, credentials to [redacted:<kind>] and terminal codes out: each part a record gives on its own before they are joined (the question, the option or action picked, the answer’s words, a decline’s reason), text as it came, and then the whole once more. A NUL and half a surrogate pair become U+FFFD. Nothing left of text is a 400 text_required, with nothing written. The bound comes last: longer words are cut after the mask, with …, never through a surrogate pair. Line endings in the words (CR, CRLF, U+2028, U+2029) are stored as LF. Candidates are listed with GET /v1/review-items?kind=memory_rule.
Authorizations
Section titled “Authorizations ”Request Body required
Section titled “Request Body required ”The body of POST /v1/memory/candidates (#2198). text is the rule in the proposer’s words; omitted, it is made from the record - the question and its answer or the action picked, what was declined and why, or the lesson - except from an answer to an access question, which is never copied (400 text_required).
object
Where a candidate rule came from (#2198), and its identity: one source is one candidate and one rule. answer: a person’s answer to a question; decline: the reason a person rejected a queue entry with; failure_choice (#2195): the action a person picked in a question on a failed test; report_lesson (#2193): a lesson of an approved mission report.
object
The record: a review item; for report_lesson, <the report’s review item>:<the lesson’s questionId>.
Trimmed and masked as the route says: with nothing left, 400 text_required.
Examples
The words made from an answered question
{ "source": { "kind": "answer", "id": "6f1c2b3a-0000-4000-8000-0000000021a9" }}A decline, in the proposer’s words
{ "source": { "kind": "decline", "id": "6f1c2b3a-0000-4000-8000-0000000000b1" }, "text": "The admin panel is not tested in this project."}The words made from the action a person picked on a failed test
{ "source": { "kind": "failure_choice", "id": "6f1c2b3a-0000-4000-8000-0000000021f5" }}The words made from a lesson of an approved report
{ "source": { "kind": "report_lesson", "id": "6f1c2b3a-0000-4000-8000-0000000021b3:6f1c2b3a-0000-4000-8000-0000000021a9" }}Responses
Section titled “ Responses ”The same source was proposed before: the entry already there, untouched
One entry of the review queue. Always starts pending; a reviewer moves it on.
object
Present on an entry disputing a RESULT. Exactly one of resultId, proposedCase, discoveredTest, question, proposedEndpoint, strategy, report, memoryRule and automation is present: an entry with two asks two questions at once, and one with none asks nothing.
Present on an entry proposing a new case. Mutually exclusive with the other kinds.
object
Content-derived identity; the dedup key.
object
What the pre-review checks found, written before the proposal was queued (#1289). It marks the proposal. Only placeholder stops anything: a proposal whose title, steps or expected result still holds {{...}}, [TODO]/[TBD]/[FIXME]/[XXX] (or the same in <...>) or a run of three or more underscores cannot be approved (409 naming the places), and that check is read off the reviewer’s edited version when there is one. Absent means nothing was found.
object
Where the rule is explained.
object
Absent means never executed — NOT “no result yet”. blocked: no answer from the target.
What a match resolves through. Absent is a normal proposal, not a degraded one: an agent that names no key gets an entry with no comparison, which is exactly what a genuinely new case gets.
An identifier another tool already uses for this test, so a report can find its case without anyone typing a Plune id. The order of kind is load-bearing: it is the ranking that decides which key wins when several match different cases.
object
The reviewer’s corrected version of the proposal (#1878), beside proposedCase, which stays what the agent said. Present only after an edit; approving the item makes the case from it. Left off the page of the listing (editedBy and editedAt stay).
object
The user id of whoever saved the version in editedCase.
When editedCase was saved.
Present on an entry reporting a test the caller RUNS and this project has no case for. Mutually exclusive with the other kinds.
object
The identity. Order does not matter — the server ranks them and the highest-ranked kind is what deduplication reads, so two adapters sending the same keys in a different order do not produce two entries about one test.
An identifier another tool already uses for this test, so a report can find its case without anyone typing a Plune id. The order of kind is load-bearing: it is the ranking that decides which key wins when several match different cases.
object
The reporter that found it, e.g. playwright.
Where the test lives, e.g. e2e/checkout.spec.ts:12. Required while everything descriptive is optional: an approved discovery becomes an automated case, and that shape needs a spec to point at.
What it did on this run, in the reporter’ own word.
The run that turned it up, once stored.
Present on an entry asking a person a question (#2189). Mutually exclusive with the other kinds.
object
Trimmed. On failure, the explanation: at most 500 characters.
The mission the question is about (#2188), one of this project’s: another project’s, or one nobody made, is a 404 mission not found.
plain only. Each trimmed; no two the same.
access only, and required there; each kind once. openapi: an API description (OpenAPI or Postman); repo: read-only access to the code; staging: a staging host with a test account; har: a HAR file.
failure only, and required there: a result of this project in a live run, failed or broken, of a live case. Another project’s, or one nobody made, is a 404; one that did not fail is a 409 result_not_failed; a second waiting question on one result is a 409 already_asked. GET /v1/review-items?resultId= finds the question for the result’s page.
failure only: the case of the result, read off the result by the server - never sent (a body that names it is a 400).
failure only, and required there: why the test failed. app-bug the application; selector, timing, data, environment, dependency; and Cairn’s covered-by-overlay, ambiguous-locator, dead-click. Copied onto the answer.
failure only, and required there: one to three actions, each kind once. A page of the queue carries a fix without its code (codeGiven: true).
An action a failure question offers (#2195), carried out by the answer that picks it (answer.action), as the person who answered. file-bug: a bug from the result, titled title, its case linked and the result its evidence - or, with bugId, the result added as evidence to that bug of the project (one of the two). accept-fix: the case’s code (codeSnapshot) becomes code; automated cases only (else 400 action_not_for_case). expected-change: the product changed on purpose - a manual case’s expected result becomes expected, an automated case’s code becomes code. quarantine: the case is muted, and its results in open runs count as muted. baseHash (sha256 hex of the code the fix was written against): when given and the case’s code has moved since, the answer is a 409 code_changed.
object
file-bug: the title of the bug to file.
file-bug: a live bug of the project the result is added to; another one is a 404.
accept-fix, expected-change: the whole fixed test() call, at most 16 384 bytes.
expected-change on a manual case: its new expected result.
On a page of the queue, in place of code.
Present on an entry proposing an endpoint for the project’s API catalog (#2192); approving it writes the endpoint. Mutually exclusive with the other kinds.
object
The API it is for; left out, the project’s only one (or none yet). With several APIs it is named (400 api_required)
Upper or lower case; stored upper case
A path TEMPLATE: it starts with /, a parameter is a whole segment written {name}, and a segment that names one thing - digits, a UUID, a long hex run, anything with @ - is refused: write it as a parameter. No query, no fragment.
derived when left out; approving says the endpoint belongs in the catalog, not that its description is right
The auth scheme names it asks for, as a spec names them; [] is public
Where the agent saw it
When the agent saw it; a time ahead of the server is taken as the moment of the approval
Present on an automation proposal (#2201, ADR 0080) - code an agent wrote for a case; a person approving it in a session writes the case’s code snapshot. Mutually exclusive with the other kinds.
object
object
At most 16 384 bytes.
object
object
What a RUN cost — never what one proposal cost. Both numbers are required together because one alone says nothing: a token count is a fact only once you know how many cases it bought. Zero is a report; a run that did not count omits the object entirely.
object
The person’s answer to question, once given: an answered question is approved, and an approved question always has one.
object
Trimmed.
The index into question.options.
What the person gives, each kind once, all of them among question.access.
#2195: the action of a failure question the person picks; this answer carries it out.
The user id of the person who answered.
failure (#2195): the question’s category, beside the chosen action - what project memory reads.
failure, with an action: what it did - the bug filed or added to (bugId), or the case changed (testCaseId).
object
#2195: an answer whose action is being carried out right now, on a question that is still pending. While it runs, any PATCH and another answer are a 409 answer_in_progress. Gone once the answer settles (the question approved, with answer) or is refused (the question waits as before). The hold is renewed while the action runs; one not renewed for 5 minutes, its process taken to have died, is not shown and holds nothing. With outcome it never runs out: the action is committed and does not run again, and the next answer ends the question from it - 200 for the same action; another answer or a rejection ends it too and is a 409 already_decided. Approving it is a 409 answer_required, and a note or any other PATCH is a 409 answer_in_progress: neither ends it. It takes its place under the queue’s ceiling until one of them comes.
object
Who is answering.
When the answer began.
The action being carried out.
The action is done, and only settling the question is missing: what it did, as answer.outcome will say.
object
Present on an entry proposing a mission’s test strategy (#2190), or the project’s test plan (projectId, #2741). Mutually exclusive with the other kinds. Approving it makes the next version of that strategy or plan.
object
The mission the strategy is for
#2741: the project whose test plan this is
A test plan in the sections of ISTQB CTFL v4.0.1 (ADR 0074).
object
What the testing is for
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
object
What is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
What is deliberately NOT tested, and why - required
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
How it is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
The methodology profile of the project (ADR 0074 §2)
object
Test levels, CTFL v4.0.1 §2.2.1
Quality characteristics, ISO/IEC 25010:2023
Test techniques, CTFL v4.0.1 §4.2-§4.4
Per cent by API, UI and by hand; 100 in all
object
Why these levels, characteristics and techniques
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A product risk: its likelihood and impact (the level is their product, 1-9) and the cases and requirements it links to.
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Live cases of the project that cover it (checked)
Requirements of the project it is about (checked); their cases are in the matrix
When the testing can start
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
When it is done
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A mission’s proposal only: the version of the project’s test plan the plan was taken from (POST /v1/missions/{id}/strategy/from-test-plan, #2740); absent for a written plan
A person’s corrected plan (#2190, PUT /v1/review-items/{id}/strategy), beside strategy, which stays what the agent said; approving takes it. editedBy / editedAt say who and when.
object
What the testing is for
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
object
What is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
What is deliberately NOT tested, and why - required
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
How it is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
The methodology profile of the project (ADR 0074 §2)
object
Test levels, CTFL v4.0.1 §2.2.1
Quality characteristics, ISO/IEC 25010:2023
Test techniques, CTFL v4.0.1 §4.2-§4.4
Per cent by API, UI and by hand; 100 in all
object
Why these levels, characteristics and techniques
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A product risk: its likelihood and impact (the level is their product, 1-9) and the cases and requirements it links to.
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Live cases of the project that cover it (checked)
Requirements of the project it is about (checked); their cases are in the matrix
When the testing can start
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
When it is done
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Present on an entry carrying a mission’s completion report (#2193), as the server built it. Mutually exclusive with the other kinds. Approving it makes the next version of the mission’s report.
object
When the mission began: results and bugs count from here
object
Risks with a case executed since the mission began
object
In the strategy’s risks
object
The latest result since the mission began; null: not run since
object
object
object
The latest result since the mission began; null: not run since
Not reported: Cairn does not send them to the platform yet
object
Not reported: Cairn does not send them to the platform yet
object
object
What the strategy deliberately left out, and why
object
object
object
object
Answered plain questions of the mission - candidates for the project’s memory (E7)
object
changes_requested (#1276): a reviewer asked the author for changes (changeRequest says what). It is the item’s own state, not a mark on a reviewer: it cannot be approved until the author re-submits, which puts the item back to pending.
Raised by every re-submission; absent is 1. An approval counts only for the revision it was given on.
On GET /v1/review-items/{id} only (#2200): the caller put the entry in the queue. The review policy refuses the author some decisions (approving without self-merge, asking for changes). Absent for anyone else and on the list.
The user ids asked to review the item - the project’s default reviewer, when they could be asked.
The default reviewer who was NOT asked because they no longer have the right to review - the list says who was passed over.
Who approved which revision; the policy’s approvalsRequired counts those of the current revision.
object
The open request for changes (present with state: changes_requested); the author’s re-submission clears it.
object
Present on every item rejected after R4, and absent forever on rejections stored before it — those are never migrated and must keep reading without error. An item that is pending or approved usually has none, but MAY still carry one: a rejection returned to review keeps the pair in the row as a draft, and the next move to rejected must supply it in the request anyway. Read it as “meaningful only while state is rejected”, not as “absent otherwise”.
Example
{ "schemaVersion": 1, "id": "6f1c2b3a-0000-4000-8000-0000000021c8", "memoryRule": { "text": "Which flow is the checkout: the one-page form or the three steps?\nAnswer: Three steps", "source": { "kind": "answer", "id": "6f1c2b3a-0000-4000-8000-0000000021a9" } }, "state": "pending", "createdAt": "2026-09-16T09:12:04.151Z", "updatedAt": "2026-09-16T09:12:04.151Z"}Proposed, pending
One entry of the review queue. Always starts pending; a reviewer moves it on.
object
Present on an entry disputing a RESULT. Exactly one of resultId, proposedCase, discoveredTest, question, proposedEndpoint, strategy, report, memoryRule and automation is present: an entry with two asks two questions at once, and one with none asks nothing.
Present on an entry proposing a new case. Mutually exclusive with the other kinds.
object
Content-derived identity; the dedup key.
object
What the pre-review checks found, written before the proposal was queued (#1289). It marks the proposal. Only placeholder stops anything: a proposal whose title, steps or expected result still holds {{...}}, [TODO]/[TBD]/[FIXME]/[XXX] (or the same in <...>) or a run of three or more underscores cannot be approved (409 naming the places), and that check is read off the reviewer’s edited version when there is one. Absent means nothing was found.
object
Where the rule is explained.
object
Absent means never executed — NOT “no result yet”. blocked: no answer from the target.
What a match resolves through. Absent is a normal proposal, not a degraded one: an agent that names no key gets an entry with no comparison, which is exactly what a genuinely new case gets.
An identifier another tool already uses for this test, so a report can find its case without anyone typing a Plune id. The order of kind is load-bearing: it is the ranking that decides which key wins when several match different cases.
object
The reviewer’s corrected version of the proposal (#1878), beside proposedCase, which stays what the agent said. Present only after an edit; approving the item makes the case from it. Left off the page of the listing (editedBy and editedAt stay).
object
The user id of whoever saved the version in editedCase.
When editedCase was saved.
Present on an entry reporting a test the caller RUNS and this project has no case for. Mutually exclusive with the other kinds.
object
The identity. Order does not matter — the server ranks them and the highest-ranked kind is what deduplication reads, so two adapters sending the same keys in a different order do not produce two entries about one test.
An identifier another tool already uses for this test, so a report can find its case without anyone typing a Plune id. The order of kind is load-bearing: it is the ranking that decides which key wins when several match different cases.
object
The reporter that found it, e.g. playwright.
Where the test lives, e.g. e2e/checkout.spec.ts:12. Required while everything descriptive is optional: an approved discovery becomes an automated case, and that shape needs a spec to point at.
What it did on this run, in the reporter’ own word.
The run that turned it up, once stored.
Present on an entry asking a person a question (#2189). Mutually exclusive with the other kinds.
object
Trimmed. On failure, the explanation: at most 500 characters.
The mission the question is about (#2188), one of this project’s: another project’s, or one nobody made, is a 404 mission not found.
plain only. Each trimmed; no two the same.
access only, and required there; each kind once. openapi: an API description (OpenAPI or Postman); repo: read-only access to the code; staging: a staging host with a test account; har: a HAR file.
failure only, and required there: a result of this project in a live run, failed or broken, of a live case. Another project’s, or one nobody made, is a 404; one that did not fail is a 409 result_not_failed; a second waiting question on one result is a 409 already_asked. GET /v1/review-items?resultId= finds the question for the result’s page.
failure only: the case of the result, read off the result by the server - never sent (a body that names it is a 400).
failure only, and required there: why the test failed. app-bug the application; selector, timing, data, environment, dependency; and Cairn’s covered-by-overlay, ambiguous-locator, dead-click. Copied onto the answer.
failure only, and required there: one to three actions, each kind once. A page of the queue carries a fix without its code (codeGiven: true).
An action a failure question offers (#2195), carried out by the answer that picks it (answer.action), as the person who answered. file-bug: a bug from the result, titled title, its case linked and the result its evidence - or, with bugId, the result added as evidence to that bug of the project (one of the two). accept-fix: the case’s code (codeSnapshot) becomes code; automated cases only (else 400 action_not_for_case). expected-change: the product changed on purpose - a manual case’s expected result becomes expected, an automated case’s code becomes code. quarantine: the case is muted, and its results in open runs count as muted. baseHash (sha256 hex of the code the fix was written against): when given and the case’s code has moved since, the answer is a 409 code_changed.
object
file-bug: the title of the bug to file.
file-bug: a live bug of the project the result is added to; another one is a 404.
accept-fix, expected-change: the whole fixed test() call, at most 16 384 bytes.
expected-change on a manual case: its new expected result.
On a page of the queue, in place of code.
Present on an entry proposing an endpoint for the project’s API catalog (#2192); approving it writes the endpoint. Mutually exclusive with the other kinds.
object
The API it is for; left out, the project’s only one (or none yet). With several APIs it is named (400 api_required)
Upper or lower case; stored upper case
A path TEMPLATE: it starts with /, a parameter is a whole segment written {name}, and a segment that names one thing - digits, a UUID, a long hex run, anything with @ - is refused: write it as a parameter. No query, no fragment.
derived when left out; approving says the endpoint belongs in the catalog, not that its description is right
The auth scheme names it asks for, as a spec names them; [] is public
Where the agent saw it
When the agent saw it; a time ahead of the server is taken as the moment of the approval
Present on an automation proposal (#2201, ADR 0080) - code an agent wrote for a case; a person approving it in a session writes the case’s code snapshot. Mutually exclusive with the other kinds.
object
object
At most 16 384 bytes.
object
object
What a RUN cost — never what one proposal cost. Both numbers are required together because one alone says nothing: a token count is a fact only once you know how many cases it bought. Zero is a report; a run that did not count omits the object entirely.
object
The person’s answer to question, once given: an answered question is approved, and an approved question always has one.
object
Trimmed.
The index into question.options.
What the person gives, each kind once, all of them among question.access.
#2195: the action of a failure question the person picks; this answer carries it out.
The user id of the person who answered.
failure (#2195): the question’s category, beside the chosen action - what project memory reads.
failure, with an action: what it did - the bug filed or added to (bugId), or the case changed (testCaseId).
object
#2195: an answer whose action is being carried out right now, on a question that is still pending. While it runs, any PATCH and another answer are a 409 answer_in_progress. Gone once the answer settles (the question approved, with answer) or is refused (the question waits as before). The hold is renewed while the action runs; one not renewed for 5 minutes, its process taken to have died, is not shown and holds nothing. With outcome it never runs out: the action is committed and does not run again, and the next answer ends the question from it - 200 for the same action; another answer or a rejection ends it too and is a 409 already_decided. Approving it is a 409 answer_required, and a note or any other PATCH is a 409 answer_in_progress: neither ends it. It takes its place under the queue’s ceiling until one of them comes.
object
Who is answering.
When the answer began.
The action being carried out.
The action is done, and only settling the question is missing: what it did, as answer.outcome will say.
object
Present on an entry proposing a mission’s test strategy (#2190), or the project’s test plan (projectId, #2741). Mutually exclusive with the other kinds. Approving it makes the next version of that strategy or plan.
object
The mission the strategy is for
#2741: the project whose test plan this is
A test plan in the sections of ISTQB CTFL v4.0.1 (ADR 0074).
object
What the testing is for
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
object
What is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
What is deliberately NOT tested, and why - required
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
How it is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
The methodology profile of the project (ADR 0074 §2)
object
Test levels, CTFL v4.0.1 §2.2.1
Quality characteristics, ISO/IEC 25010:2023
Test techniques, CTFL v4.0.1 §4.2-§4.4
Per cent by API, UI and by hand; 100 in all
object
Why these levels, characteristics and techniques
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A product risk: its likelihood and impact (the level is their product, 1-9) and the cases and requirements it links to.
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Live cases of the project that cover it (checked)
Requirements of the project it is about (checked); their cases are in the matrix
When the testing can start
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
When it is done
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A mission’s proposal only: the version of the project’s test plan the plan was taken from (POST /v1/missions/{id}/strategy/from-test-plan, #2740); absent for a written plan
A person’s corrected plan (#2190, PUT /v1/review-items/{id}/strategy), beside strategy, which stays what the agent said; approving takes it. editedBy / editedAt say who and when.
object
What the testing is for
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
object
What is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
What is deliberately NOT tested, and why - required
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
How it is tested
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
The methodology profile of the project (ADR 0074 §2)
object
Test levels, CTFL v4.0.1 §2.2.1
Quality characteristics, ISO/IEC 25010:2023
Test techniques, CTFL v4.0.1 §4.2-§4.4
Per cent by API, UI and by hand; 100 in all
object
Why these levels, characteristics and techniques
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
A product risk: its likelihood and impact (the level is their product, 1-9) and the cases and requirements it links to.
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Live cases of the project that cover it (checked)
Requirements of the project it is about (checked); their cases are in the matrix
When the testing can start
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
When it is done
One item of a strategy. Exactly one of source, assumption, waitingOn - none, or two, and the strategy does not pass its checks (strategy_lint).
object
Markdown
Where an item was read. mission-source: the id of a source of this mission; requirement: the id of a requirement of the project (both checked); doc-section (docs/checkout.md#payment), screen (SCR-03 Cart), endpoint (POST /cart), url: a place in words a person can follow.
object
«припущення, підтвердіть»: the agent assumed it and asks a person to confirm
The id of a question of this mission (a review item) whose answer this part waits for; the text says what waits, never a guessed answer
Present on an entry carrying a mission’s completion report (#2193), as the server built it. Mutually exclusive with the other kinds. Approving it makes the next version of the mission’s report.
object
When the mission began: results and bugs count from here
object
Risks with a case executed since the mission began
object
In the strategy’s risks
object
The latest result since the mission began; null: not run since
object
object
object
The latest result since the mission began; null: not run since
Not reported: Cairn does not send them to the platform yet
object
Not reported: Cairn does not send them to the platform yet
object
object
What the strategy deliberately left out, and why
object
object
object
object
Answered plain questions of the mission - candidates for the project’s memory (E7)
object
changes_requested (#1276): a reviewer asked the author for changes (changeRequest says what). It is the item’s own state, not a mark on a reviewer: it cannot be approved until the author re-submits, which puts the item back to pending.
Raised by every re-submission; absent is 1. An approval counts only for the revision it was given on.
On GET /v1/review-items/{id} only (#2200): the caller put the entry in the queue. The review policy refuses the author some decisions (approving without self-merge, asking for changes). Absent for anyone else and on the list.
The user ids asked to review the item - the project’s default reviewer, when they could be asked.
The default reviewer who was NOT asked because they no longer have the right to review - the list says who was passed over.
Who approved which revision; the policy’s approvalsRequired counts those of the current revision.
object
The open request for changes (present with state: changes_requested); the author’s re-submission clears it.
object
Present on every item rejected after R4, and absent forever on rejections stored before it — those are never migrated and must keep reading without error. An item that is pending or approved usually has none, but MAY still carry one: a rejection returned to review keeps the pair in the row as a draft, and the next move to rejected must supply it in the request anyway. Read it as “meaningful only while state is rejected”, not as “absent otherwise”.
Example
{ "schemaVersion": 1, "id": "6f1c2b3a-0000-4000-8000-0000000021c8", "memoryRule": { "text": "Which flow is the checkout: the one-page form or the three steps?\nAnswer: Three steps", "source": { "kind": "answer", "id": "6f1c2b3a-0000-4000-8000-0000000021a9" } }, "state": "pending", "createdAt": "2026-09-16T09:12:04.151Z", "updatedAt": "2026-09-16T09:12:04.151Z"}A body that is not a candidate, or code: text_required: an answer to an access question without text, or a text with nothing left once it is masked
object
Example
{ "error": "an answer to an access question is not copied into a rule - say the rule in `text`", "code": "text_required"}No / invalid token or session
A reader, or a read token
object
Example generated
{ "error": "example"}The source is not an item of this project (for report_lesson, the part before the colon)
object
Example generated
{ "error": "example"}code: not_a_source: the item is no answered question, or one answered with an action (answer); no entry rejected with a reason (decline) - a question and a candidate rule are not declines; no question on a failed test answered with an action (failure_choice); no approved report holding that lesson (report_lesson)
object
Example
{ "error": "an answer is a source once a person answered the question - this item is no answered question", "code": "not_a_source"}Two refusals share this status, and what to do about them is opposite. Rate-limited — the per-route throttle, or the per-minute request limit of the plan an API token counts by (#1442, #2092): Retry-After carries the seconds until the window rolls over, and waiting that long is the way out. Quota reached — a ceiling on what the project holds (runs in the last 24 hours, test cases, items waiting in the review queue, rules in force in the project’s memory): { error, quota }. A retry straight away frees nothing, so do not loop on it; the message says what makes room — deleting cases, working the queue down, an operator raising the limit — and the run quota, a rolling 24 hours, frees its oldest runs as they age out. Tell the two apart by the body: a 429 whose body has a quota object is a quota, one without it is a throttle. quota carries kind, limit, used, remaining and resetsAt: the figures behind the sentence, and for the run quota the instant its oldest run ages out (#1258). Retry-After is sent on a quota only where there is such an instant (the run quota: the whole seconds, at least 1, until resetsAt); the case, review-queue and project quotas free only by an action, have resetsAt: null, and send no Retry-After. GET /v1/quotas reads the same figures without a refusal.
object
The body of a 429 that is a quota and not a throttle: the sentence, and the figures behind it.
object
What was reached and the way out.
object
The ceiling. A write that would take used past it is refused with 429.
What the project holds now, counted by the same query that refuses a write.
limit - used, never below 0.
When the oldest counted run ages out of the rolling 24 hours and a slot frees on its own. null for the two quotas only an action frees (delete cases, decide queue items), and for a run quota with no runs in the window.
Examples
A throttle: no `quota`; wait `Retry-After`
{ "error": "rate limit exceeded — slow down"}A quota (#2458): the MCP server ceiling; `resetsAt` is null, nothing frees it by waiting
{ "error": "MCP server quota reached — at most 50 servers per project, and a server cannot be removed yet. Ask an operator to raise the limit.", "quota": { "kind": "mcpServersPerProject", "limit": 50, "used": 50, "remaining": 0, "resetsAt": null }}Headers
Section titled “Headers ”Seconds until the window rolls over on a throttle; on a quota, until quota.resetsAt — sent only for the run quota, which has one