feat(spec): retire RestApiEndpoint.handlerStatus and the Route Coverage Report shapes (#13823)

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code

[claude]

Fixes#13823

Ruling implemented

Director decision batch #27, maintainer 2026-09-01 verbatim 「同意」 (issue comment 5494755488), implemented exactly:

1. remove:RestApiEndpointSchema.handlerStatus 按 spec-property-retirement 手册退役(liveness 台账 verdict 确认 → ADR-0087 转换 → 墓碑 + 生成基线/表单/文档/pin 全套);RouteCoverageEntrySchema/RouteCoverageReportSchema 随同一 PR 按 liveness 读数处置(零构造者即同退);
2. 前置一扫:退役前跑教材扫描 —— 除 flight ⑦ 已纠的那句外,还有哪些 published skill / 文档 / 示例在教 handlerStatus;非零则同笔清扫(教材曾教过不存在的 501 行为,是本卡的出处);
3. enforce 排除:「声明未建的路由回 501」是零拉力的新能力,不是修复;
4. 类方向入册(finding(spec): all six branded identifier schemas in branded-types.zod.ts are declared but unbound — no schema anywhere composes them #13612 / finding(spec): EventNameSchema's only three binding schemas have no runtime consumer — the live event vocabulary is a closed enum that never touches it #13613 引用用):ADR-0049 declared-but-unenforced 的默认处置 = retire(零拉力键);enforce/bind 只当有具名消费者或实测拉力。
5. 条款②:YES(发布 accept 面收窄)⇒ draft + needs:contract-review 同笔,本席复审;changeset 按退役惯例。

Clause ② applies (the published accept set narrows): draft PR, needs:contract-review on both carriers, in-seat review. Item 4 is recorded here as the class direction for #13612 / #13613 — not ruled by this PR.

Premise re-measured on origin/main (a9b2be0, 2026-09-02)

Identifier search for handlerStatus, skills/** and tests excluded — three hits, all in the declaring file:

site	what
packages/spec/src/api/plugin-rest-api.zod.ts:177	the declaration on RestApiEndpointSchema (HandlerStatusSchema.optional())
packages/spec/src/api/plugin-rest-api.zod.ts:1415	the same enum re-declared on RouteCoverageEntrySchema
packages/spec/src/api/plugin-rest-api.zod.ts:1431	the docblock saying adapters SHOULD warn when handlerStatus !== 'implemented'

Plus CHANGELOG.md:1502 (history) and generated artifacts. DispatcherErrorCode.enum.NOT_IMPLEMENTED emitters, none consulting the key: packages/runtime/src/endpoint-executor.ts:320,505,517, packages/runtime/src/api-mapping.ts:225, and one more than the card counted — packages/runtime/src/api-endpoint-step.ts:343 — every one the declarative-endpoint executor refusing a target or mapping it cannot serve. RouteCoverageEntrySchema / RouteCoverageReportSchema constructors: zero in objectstack (only the type-alias pin and generated artifacts reference them), zero in objectui at the pinned sha d8ec8d6d, zero in cloud. RestApiEndpointSchema itself is parsed by nothing in the tree outside its own unit tests. Premise still valid; no consumer has appeared.

Teaching sweep (before retiring)

path	hit	action
content/docs/references/api/plugin-rest-api.mdx:200,375,408,424,442	generated reference rows for handlerStatus / the coverage shapes	AUTO-GEN — regenerated by gen:docs (tombstone row now reads [REMOVED] …; coverage sections gone)
content/docs/references/index.mdx:90	generated export list naming HandlerStatus, RouteCoverageEntry, RouteCoverageReport	AUTO-GEN — regenerated
content/docs/** hand-written trees	no hit for handlerStatus; the 501 prose at api/index.mdx:161,171, kernel/services-checklist.mdx:193, permissions/authentication.mdx:1243,1281 describes the service-slot / discovery status 501, not this key	not the false claim — no action
examples/**, packages/*/README.md, packages/create-objectstack/**	zero hits	none
skills/objectstack-api/SKILL.md (at this PR's base a9b2be0: :136handlerStatus: 'implemented', inside the os:check-marked RestApiEndpointSchema.parse({...}) example; :380 the 501 NOT_IMPLEMENTED — Route declared but handler is a stub table row; :383-391 the "Handler Status" section; :393-395 flight ⑦'s PR #13808 correction)	four teaching sites at the base	⛔ governed surface, not edited. None on main: bae4088f3 (#14448, 08:27Z today) cut every one — git grep -n -i -E 'handlerStatus|handler status|RouteCoverage|HandlerStatus' origin/main -- skills/ returns zero hits at 99b8faf8f, with RestApiEndpoint still hitting the skill as the live control — so no skills-lane follow-up is needed. (check:skill-examples was measured green against the base's marked block too: parse takes unknown, so the tombstone's never input type never reached it.)

Retirement steps (spec-property-retirement playbook)

• Route.RestApiEndpointSchema is a non-strict lazySchema(() => z.object(...)) ⇒ retiredKey() tombstone for the key (packages/spec/src/api/plugin-rest-api.zod.ts:208-217), prescription hoisted as HANDLER_STATUS_RETIRED (:131-138) under the retirement record (:91-130). RouteCoverageEntrySchema / RouteCoverageReportSchema ⇒ whole-def removal, route 3 (nobody parses or constructs one), recorded at the file tail (:1439-1453). HandlerStatusSchema / HandlerStatus ⇒ orphan value def once both carriers are gone (refactor(spec)!: remove the plugin sandboxing / integrity / approval config that never existed (#3896 follow-up) #3950, the ui/ThemeMode precedent) — removed with them.
• Liveness ledger. Verdict dead, confirmed by the re-measure above. No ledger row moves: RestApiEndpointSchema is outside the walked population — it is not a registered metadata type and not in SPEC_ONLY_SCHEMAS (packages/spec/scripts/liveness/check-liveness.mts:328) — so there is no row to write and adding one would be an ORPHAN. The verdict is recorded in the schema comment, the D3 entry and the changeset.
• ADR-0087. Tail-appended entries, registry.ts regenerated by gen:migration-registry: entries/retired-keys/18.api__RestApiEndpoint__handlerStatus.ts; entries/retired-defs/18.api__HandlerStatus.ts, 18.api__RouteCoverageEntry.ts, 18.api__RouteCoverageReport.ts; entries/semantic/18.rest-api-endpoint-handler-status-retired.ts. No D2 conversion, deliberately — a REST API plugin route registration is not a stack collection member and never a sys_metadata row, so a MetadataConversion would be a transform with no seam that ever runs (the kernel/Manifest:loading / kernel/KernelContext:previewMode disposition); for the same reason the tombstone carries no os migrate meta sentence. Registered under 18 (launch-window convention). spec-changes.json and the upgrade guide are unchanged because both project majors up to PROTOCOL_MAJOR (17) only — the same as every 18-registered precedent.
• Generated baselines.json-schema.manifest/api.json −3 keys (deleted deliberately; the json-schema.manifest.json 的「deliberate removal」删行仍是纪律而非门禁 —— #4650 的同类洞,上移一层(整 schema 级) #4725 manifest deletion gate adjudicated all three against RETIRED_DEFS_BY_MAJOR[18]); authorable-surface/api.json −10 rows for the two defs (route 3, check (c) proof) and api/RestApiEndpoint:handlerStatus → … [RETIRED]; api-surface/api.json, export-origins/api.json, declaration-map/api.json −6 names each; content/docs/references/** and the strictness-ledger counts regenerated by check:generated --fix. authorable-surface.base.json lags (allowed; check:authorable-surface green).
• Forms / i18n. No *.form.ts offers the key; no bundle changes.
• Pin tests.packages/spec/src/api/plugin-rest-api.handler-status-retirement.test.ts: all three former values (including the documented default 'implemented' — prose only, never a Zod .default(), so no residue window) refused with code: 'invalid_type', path: ['handlerStatus'] and the prescription text (ADR named, no issue id); the same refusal through the route-registration embed at ['endpoints', 0, 'handlerStatus']; positive control — a well-formed endpoint without the key parses and grows no handlerStatus property; the eight shipped default route registrations still parse; a tsc-channel pin (@ts-expect-error, file proven in the tsconfig.test.json program via --listFiles); zero holders for all six retired names on every public entry with the carriers surviving; RETIRED_KEYS_BY_MAJOR[18] / RETIRED_DEFS_BY_MAJOR[18] / step-18 semantic membership. type-alias-convention.pin.test.ts: pins Iso123 / Iso125 / Iso126 left with their schemas, count rebalanced 835 → 832 with the file's history note.
• Changeset..changeset/rest-api-endpoint-handler-status-retired.md — @objectstack/spec minor, BREAKING under the launch-window convention, FROM → TO + one-line fix, ADR-0087 marker registered rest-api-endpoint-handler-status-retired (check:adr-0087-registration green).
• Enforce excluded. No 501 stub mounting, no registrar change.

Verification

Every heavy step ran under scripts/pm/os-verify-lock.sh (OS_VERIFY_LOCK_SLOT=issue-13823); every exit captured before any pipe; the gates' own verdict lines quoted.

Final head 875e3fa5e (chain 4): pnpm --filter @objectstack/spec build exit 0 (gen:schema: "3 schema(s) left the published set since a9b2be0, each declared (#4725)" and "2 baseline deletion(s) since a9b2be0 carry their own proof (#4650)"; check-dts-emitted: 34/34); check:generated (check-only) exit 0 — "✓ All 15 generated artifacts are up to date."; targeted vitest via exec vitest run --maxWorkers=2 FILES (type-alias-convention.pin, plugin-rest-api.handler-status-retirement, plugin-rest-api, retired-key-migrate-sentence, migrations) — "Test Files 5 passed (5)" / "Tests 172 passed (172)"; check:skill-examples exit 0 — "✅ 265 prose examples type-check across 3 surface(s)" (the @objectstack/client-react closure was built for it); check:adr-0087-registration exit 0 — "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition"; check:empty-changeset, check-changeset-no-major, check:nul-bytes, check:doc-authoring, check:spec-parsed-alias all exit 0; git status --porcelain empty after the build (the authorable-surface.base.json anchor was not rewritten — it lags, allowed).

7cbf0ed60 (chain 3, the pin-literal fix): targeted vitest 5 files / 172 tests passed; check:test-typecheck exit 0; check:spec-parsed-alias / check:doc-authoring / check:nul-bytes exit 0; check:api-surface refused to judge ("dist/**/*.d.ts is OLDER than packages/spec/src" — the test-file edit post-dated the build; not a measurement, settled by the rebuild above).

90d567cd3 (chain 2, the full union): pnpm --filter @objectstack/spec typecheck exit 0 (tsc, the scripts tsconfig, and "check:test-typecheck: OK — 54 file(s) / 262 error(s) / 146 pinned signature(s)", ledger unchanged, so the new test file compiles clean; its membership in the tsconfig.test.json program proven with tsc --listFiles); full spec vitest--maxWorkers=2: "Test Files 1 failed | 451 passed (452)" / "Tests 1 failed | 12190 passed (12191)" — the single failure was still declares all 832 isomorphic pins (the toHaveLength literal still 835; the same red CI reported on Test Core 2/6), fixed in 7cbf0ed60 and green in chains 3 and 4; check:skill-examples exit 0; check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries exit 0; check:generated exit 0 (15/15); check:doc-authoring / check:nul-bytes / check:adr-0087-registration / check:spec-parsed-alias exit 0; check-dev-prereqs exit 1 = "The workspace is not built — 32 of 67 workspace packages declare an entry point under dist/ that is not on disk" (a pnpm dev precondition on apps and services, unrelated to this diff — NOT MEASURED, not red).

Earlier on the same tree (before the docs-only, test-literal and changeset-prose commits): check:variant-docs, check:empty-state, check:llms-txt, check:yaml-examples, check:template-manifests, check:error-code-provenance, check:objectui-pin-citations, check:quick-reference-counts, check:corpus-claim-drift, check:skill-identifier-liveness, check:doc-anchors, check:docs-single-h1, check:docs-audit-scope, check:test-source-alias, check:cross-package-test-inputs, check:published-files, check:type-check-coverage, check:migration-registry and check:liveness (both inside check:generated), and the root check-doc-frontmatter / check-docs-section-name / check-doc-route-spelling / check-section-landing-index / check-keyed-text-bounds / check-ci-filter-parity / check-comment-mask-adoption / check-shard-attestation / check-system-context-census / check-undeclared-dep-imports / check-plugin-teardown-shape / docs-audit/check-affected-docs / docs-audit/check-drift-comment / pm/check-half-states scripts — all exit 0. Earlier reds that this PR fixed before landing: check:doc-authoring (the tombstone string carried #13823; now cites ADR-0049 only, with a negative pin) and the pin-count test above.

Declared narrowing: the full 452-file spec suite ran on 90d567cd3; the two later commits change one assertion literal in a test file and changeset prose, re-verified with the targeted files rather than the whole suite (≈6 min held under a contended lock) — CI's Test Core runs the full suite on the final head. Gate list derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack (no paths; 74 commands at b862317); what was not run is the CI-owned farm named below. CI on 875e3fa at report time: 31 check-runs — completed/skipped: 2, completed/success: 14, in_progress/None: 15.

Not measured here (CI-owned)

• check:type-check-coverage --re-measure / check:type-check-debt (needs the whole package closure built — PREREQUISITE NOT MET, exit 3, not a finding), check:test-completeness (grades CI's turbo test log), check:react-declaration-parity (needs objectui's manifest), the repo-wide pnpm lint and the rest of the 74-command derivation not named above.

🤖 Generated with Claude Code

https://claude.ai/code/session_01GDA48PuRFrHyRfdkBz8m21

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

14 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 9 changed file(s) yielded no anchor (packages/spec/api-surface/api.json, packages/spec/authorable-surface/api.json, packages/spec/declaration-map/api.json, …) — pages documenting those are invisible to this run
• 10 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 47 of 219 client-bound route-ledger rows — the other 172 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 172: 14 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 56 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 102 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.

Coarse fallback — 128 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from ba96272f6fedf9a4afda784a9aaaf3e413d04262 — the merge of head 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 into base 99b8faf8f937d8d2992f1bf167e418687c5f2f71, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin ba96272f6fedf9a4afda784a9aaaf3e413d04262 && git checkout ba96272f6fedf9a4afda784a9aaaf3e413d04262
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 99b8faf8f937d8d2992f1bf167e418687c5f2f71 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5 && git checkout -B drift-repro 99b8faf8f937d8d2992f1bf167e418687c5f2f71 && git merge --no-ff 875e3fa5eb9d9baa0c644041e001cb1148f3f2f5
node scripts/docs-audit/affected-docs.mjs --json 99b8faf8f937d8d2992f1bf167e418687c5f2f71

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

[claude]

CI red on head 90d567cd — this PR's own; patch round requested from the dispatched dev. (domain:spec seat, session session_01GDA48PuRFrHyRfdkBz8m21, seat post #6017.)

• Failing check: Test Core (2/6) — packages/spec/src/type-alias-convention.pin.test.ts > "still declares all 832 isomorphic pins": expected [ Array(832) ] to have a length of 835 but got 832 at line 2093.
• Cause: the diff updated the case title, the section header and added the 835 -> 832 receipt prose, but the assertion literal expect(pins).toHaveLength(835) directly under the receipt still reads 835. One-literal fix (835 -> 832); every other file in that shard passed (450 passed / 1 failed).
• Not a base failure: the same case is green on origin/main at 835 pins; the count moved because this PR removed Iso123 / Iso125 / Iso126 with the retired defs, which is the intended shape.
• Action: the dev is fixing the literal on this branch as a new commit (no history rewrite), re-running the spec test locally, then pushing. Contract review resumes on the pushed head; the landing gate stays "every check on the current head green".

---

Generated by Claude Code