spec(ui): declare titleField on KanbanConfigSchema — the optional key its five item-titled siblings already carry

[os-bill]

Fixes #16894

Clause-②: yes (widening)

KanbanConfigSchema now declares titleField as optional z.string(), executing the director seat's decision batch #87 (recorded at objectstack-ai/objectui#8367 comment 5582071618, confirmed by the maintainer verbatim 「批 #87 同意」). The direction was settled there; this PR is the implementation only.

What moved

file	change
packages/spec/src/ui/view.zod.ts	titleField: z.string().optional() on KanbanConfigSchema, plus the docblock that records why it is optional
packages/spec/src/ui/view.test.ts	the card's four-leg probe (both controls firing on the same call shape) and the optionality leg
packages/spec/authorable-surface/ui.json	regenerated — one added row, ui/KanbanConfig:titleField
content/docs/references/ui/view.mdx, content/docs/references/api/protocol.mdx, content/docs/references/data/object.mdx	regenerated by gen:docs
.changeset/16894-kanban-config-titlefield.md	@objectstack/spec: minor

Regenerated, never hand-edited: check:authorable-surface wrote the first and pnpm --filter @objectstack/spec gen:docs the rest. check:generated reports all 15 artifacts up to date.

Premise re-derivation — the card's sibling table does NOT match the tree

Re-derived on this worktree at origin/main 79a046f8c (the dispatch's own derivation ran one commit behind, at 582d3e5). git grep -n 'titleField' -- packages/spec/src/ui/view.zod.ts returns six carriers; mapping each to the schema whose member list encloses it:

line (pre-change)	owning schema	arity
:1057	GalleryConfigSchema (declared :1050)	optional
:1071	TimelineConfigSchema (declared :1065)	required
:1407	CalendarConfigSchema (declared :1401)	optional — the ADR-0079 fallback docblock the ruling names
:1490	GanttConfigSchema (declared :1484)	required
:1638	ListMapConfigSchema (declared :1631)	optional
—	KanbanConfigSchema (declared :1346)	absent — the defect

Two corrections to the card body, neither of which disturbs the ruling:

• The card's table calls Timeline optional. It is required on the tree.
• The card's table lists five schemas. There are six carriers: ListMapConfigSchema also declares titleField, optional, and the card does not mention it.

The ruling is unaffected: it prescribes the arity directly (optional z.string()) and names CalendarConfigSchema's docblock as the reference, which is optional at :1407. So the shape landed here is the ruled one, and the card's optional/required split was simply not re-measured when it was written. The defect itself re-derives exactly as filed: KanbanConfigSchema is a strictObject and refused titleField by name.

Evidence

leg	command	verdict
build	pnpm --filter @objectstack/spec build	VERDICT command-exit 0
② typecheck	pnpm --filter @objectstack/spec typecheck	TYPECHECK_EXIT=0 — check:test-typecheck: OK
② tests	pnpm --filter @objectstack/spec test	TEST_EXIT=0 — Test Files 483 passed (483) / Tests 13775 passed (13775)
targeted	pnpm --filter @objectstack/spec exec vitest run --maxWorkers=2 src/ui/view.test.ts	Test Files 1 passed (1) / Tests 376 passed (376)
artifacts	pnpm --filter @objectstack/spec check:generated	15 of 15 up to date

① is empty by construction: packages/spec has no workspace dependencies, so --filter '@objectstack/spec^...' build has an empty closure. The package itself was built before any gate that reads dist/.

③ Gate families were derived on this worktree, not inherited: node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --commands over the real 7-path change set yields 108 families. All 108 were run with each exit code landed to disk before being read, and reconciled with --ran:

dispatch-gates --ran: 108 derived family(ies) accounted for — 105 run, 3 NOT-MEASURED (3 DERIVED from a recorded exit 3)

103 green. The five non-zero results, none of them a finding against this diff:

• pnpm check:dual-build-cjs-loads, pnpm check:lean-entry-closure, pnpm check:type-check-debt — exit 3, PREREQUISITE NOT MET, all three refusing because the whole monorepo is not built on this worktree. NOT MEASURED, declared to CI, which checks out and builds fresh.
• node scripts/check-plugin-teardown-shape.mjs --self-test — exit 1, but a prerequisite refusal in its own words: "cannot read the positive control at 621a487607881c66b2899b7e3477115229a156b4 … Deepen the clone". This container's checkout is shallow. NOT MEASURED. The gate itself (without --self-test) ran green.
• pnpm check:cross-package-test-inputs — exit 1, pre-existing on the base tree, proven by control rather than asserted. With all seven changed paths restored to 79a046f8c (and the changeset removed), the gate fails identically; the finding it prints names packages/cli/test/init-created-files-summary.e2e.test.ts descending packages/spec/dist/, and this diff touches neither that test, nor turbo.json, nor any declaration table. Restore was proven by git diff HEAD empty plus a git hash-object match against every HEAD blob.

Lint, as a proven narrowing rather than a repo-wide sweep. pnpm exec eslint --no-inline-config --format json over the two changed TypeScript files: exit 0, 2 files, 0 errors, 0 warnings, the file count read from the JSON output's own length. The population it narrows from is 6798 files — computed from ESLint's own resolved config via ESLint#isPathIgnored over git ls-files, not estimated. The narrowing excludes nothing, and that is a property of this repository's config rather than a hope: eslint.config.mjs states it in its own comment — this repo "runs one eslint.config.mjs, which never enables type-aware linting (no parserOptions.project, no typed @typescript-eslint rules) for ANY file, test or not." With no rule reading types across a file boundary, a diff confined to these two files cannot move the verdict on any of the other 6796. The repo-wide pnpm lint run remains CI's.

Every reading above was taken at 6d01b4b, this branch's final commit, on a tree whose git status --porcelain is empty. origin/main had not moved from the branch point (79a046f8c) when the PR was opened, so no merge was owed.

Acceptance notes

• packages/lint does not field-check the key this PR declares. POSITIONS.kanban in packages/lint/src/validate-list-view-field-refs.ts lists groupByField (error), summarizeField (warning) and columns (warning) — and no titleField. Every sibling face lists it: calendar.titleField warns, gantt.titleField and timeline.titleField error. So from this release a stale or misspelled kanban.titleField names nothing and is reported by nothing, while the identical typo one block away is caught. Read-only on this card by dispatch, so it is reported and not edited; nothing in this PR turns it red, and no gate reconciles that table against the spec's member lists. Worth its own card.
• The node-level half (finding(types,plugin-kanban): the ruled KanbanSchema carries three zero-read members (allowCollapse, cardTemplates, columnWidths) and the board reads an undeclared titleField — enforce-or-remove on the shape objectui#7664 declared objectui#7742) has no home here, and the card's conditional resolves to "no". ObjectKanbanPropsSchema in packages/spec/src/ui/component.zod.ts already declares titleField: z.string().optional(), described as "Legacy fallback for cardTitle (the board reads cardTitle || titleField). Prefer cardTitle". That is a different semantic on a different schema in a different file — a deprecated alias for cardTitle, not a view-config field binding — so one declaration cannot serve both faces and this card is not widened. That file is also held by spec(ui): ComponentPropsMap has no rows for object-map, object-gantt and object-tree — add them from the renderers' read points (#7751 method) so 「以协议为准」 resolves for all five ladder blocks (objectui#8348 Q1, ruled C) #18305 / PR feat(spec): object-map, object-gantt and object-tree get ComponentPropsMap rows, derived from the renderers' read points (objectui#8348 Q1-C) #18403; it was read, never touched.
• Noted, not filed: KanbanConfigSchema carries no object-level .describe(), where GalleryConfigSchema, TimelineConfigSchema, CalendarConfigSchema and ListMapConfigSchema all do, so its nested-shape heading in the generated reference renders with no description sentence. Cosmetic, generated-docs only, no accept set involved — left for whichever card next edits this schema.
• The issue body was checked for sanitizer truncation: none found. It is 6732 characters through the REST read and terminates on its own Refs: line.

File surface

Two files beyond the dispatch's declared surface, both generated by the mandated gen:docs run and neither hand-edited: content/docs/references/api/protocol.mdx and content/docs/references/data/object.mdx. The kanban config shape is inlined on those two reference pages as well as on ui/view.mdx, so each picks up titleField?: string in its kanban row. Omitting them would leave check:docs red. Every hunk in all three files is this one key and nothing else.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

1 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️ 1 changed file(s) yielded no anchor (packages/spec/authorable-surface/ui.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

• 1 changed file(s) yielded no anchor (packages/spec/authorable-surface/ui.json) — pages documenting those are invisible to this run
• the SDK route bridge reached 60 of 215 client-bound route-ledger rows — the other 155 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 155: 0 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 55 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 100 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 — 136 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 79a046f8cdf085d95200826ee9bb2fa6584bc3d5 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from 67d8299911aca37ad3db044fbe890779d35791c7 — the merge of head 6d01b4b63e07b5fbdc0069ce8f5143167b9813ff into base 79a046f8cdf085d95200826ee9bb2fa6584bc3d5, 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 67d8299911aca37ad3db044fbe890779d35791c7 && git checkout 67d8299911aca37ad3db044fbe890779d35791c7
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 79a046f8cdf085d95200826ee9bb2fa6584bc3d5 6d01b4b63e07b5fbdc0069ce8f5143167b9813ff && git checkout -B drift-repro 79a046f8cdf085d95200826ee9bb2fa6584bc3d5 && git merge --no-ff 6d01b4b63e07b5fbdc0069ce8f5143167b9813ff

node scripts/docs-audit/affected-docs.mjs --json 79a046f8cdf085d95200826ee9bb2fa6584bc3d5

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

[os-bill]

Contract review

Served-tier: 30/30 CONTRACT_REVIEW_TIER
Head-sha: 6d01b4b63e07b5fbdc0069ce8f5143167b9813ff

Isolated, at-tier clause-② review of PR #18561 for card #16894, executing the director ruling (decision batch #87, objectstack-ai/objectui#8367 comment 5582071618, maintainer 「批 #87 同意」). The dispatch order and the dispatching seat's acceptance were withheld from this seat; every reading below was taken from the tree and the diff, on a detached worktree of the head, with the objectui pin 53ded82b fetched for the renderer readings. Live re-read: the PR head is still 6d01b4b6, two commits on base 79a046f8; origin/main has moved 16 commits since that base and none of them touches any of the 7 changed paths (git diff --stat 79a046f8..origin/main over the seven paths is empty), so the head reviewed is the head that would merge.

① Derived judgments

Each accept-set or public-surface change the diff produces, named and judged:

1. KanbanConfigSchema gains titleField: z.string().optional() — packages/spec/src/ui/view.zod.ts:1370 at the head. Arity matches the ruled shape (optional, string, absence falls through). Judged RIGHT. Sibling table re-derived on the head, not quoted: Gallery :1057 optional, Timeline :1071 REQUIRED, Calendar :1426 optional, Gantt :1509 required, ListMap :1657 optional. The card's table (Timeline optional, five carriers) is wrong on both counts and the PR's correction is accurate; the ruling names the arity directly, so nothing in it is disturbed.
2. Transitive accept sets: ListView.kanban, ObjectListView.kanban and the published metadata item body (api/protocol.mdx) widen by the same one key through schema reuse — one declaration, no second one anywhere. Judged RIGHT.
3. packages/spec/authorable-surface/ui.json: exactly one added row, ui/KanbanConfig:titleField (file 1218 → 1219 lines; ui/KanbanConfig:* rows 3 → 4). No other shard, no authorable-surface.base.json re-anchor, no api-surface/ or api-surface-signatures.json movement (the latter hashes define* helpers only). Re-derived: pnpm --filter @objectstack/spec build then check:generated on the head worktree → 15 of 15 artifacts up to date, check:authorable-surface, check:api-surface, check:docs, check:liveness all ✓. Judged RIGHT, and regenerated rather than hand-edited.
4. Generated reference pages: ui/view.mdx (+3 table rows in the KanbanConfig, ListView.kanban, ObjectListView.kanban nested-shape tables; 6 inline-shape lines rewritten), api/protocol.mdx (2 inline lines), data/object.mdx (1 inline line). Every hunk is this one key and nothing else — no smuggled second change. Judged RIGHT.
5. Mechanical widening control: node scripts/pm/check-widening-tells.mjs --declaration yes --diff pr.diff → exit 0 (a yes is never blocked). The control run with --declaration no → exit 4 with exactly ONE tell, T1 at packages/spec/src/ui/view.zod.ts:1370, and no T2/T3/T4 — the diff carries one widening shape, and it is the declared one.
6. The .describe() — "Field displayed as the card title. Omit to fall back to the record display name (ADR-0079 resolver chain)" — measured against the pinned renderer rather than assumed. The key IS read: plugin-list/src/ListView.tsx:1464 projects kanban.titleField into the fetch, and plugin-kanban/src/ObjectKanban.tsx:302-341 tries the explicit field first and, when it yields nothing, calls getRecordDisplayName — the ADR-0079 resolver (core/src/utils/record-title.ts:28-40: nameField → displayNameField → titleFormat → type-aware derivation → Record #id). So a layer delivers the promised fallback. ⚠️ One rung sits ahead of it at the pin: app-shell/src/views/ObjectView.tsx:2388 floors an omitted viewDef.kanban?.titleField at the literal 'name' before the node sees it, pinned by ObjectView.kanbanLane-8193.test.tsx:129 ("floors titleField at name"). objectui#7029 deleted exactly that rung for calendar, which is why the identical Calendar describe is exact and this one is exact only for records with no non-empty name value. The wording is the ruled wording (「exactly as the CalendarConfigSchema docblock already states」) and the residual rung is renderer-side in the sibling repo, so this does not refuse the spec change; it is escalated in ③ (d). Judged RIGHT as ruled, with the rung named.
7. Tests (view.test.ts:190-223): the four-leg probe — CONTROL-1 bogus key refused on the named surface, CONTROL-2 canonical accepted, PROBE titleField accepted AND survives the parse as a member — and the optionality leg (key absent on output when omitted). Mirrors the card's executable acceptance criterion. Re-run here: vitest run src/ui/view.test.ts src/ui/view-metadata-schema.test.ts → 426 passed. eslint on the two changed TS files → exit 0. Judged RIGHT.
8. Nothing else widens: no enum/union member, no export, no registry row, no ADR text (the ruling forbids one), no governed surface in the file list.

② Semver level

.changeset/16894-kanban-config-titlefield.md → "@objectstack/spec": minor, carrying Clause-②: yes (widening) — the same declaration as the PR body line 3 and the card. Right level: a new optional key on a published strict accept set is a widening, yes takes at least minor, nothing admitted before is refused, nothing is renamed or retired, so no migration text, no tombstone and no ADR-0087 marker is owed. Gates run on the head tree: check-empty-changeset ✓ (1 declaring changeset added, none from the base modified), check-adr-0087-registration ✓ (non-breaking), check-changeset-no-major ✓. The changeset text matches the diff: the six-carrier sentence and the "Timeline and Gantt spell it required" sentence check against the head lines in ① item 1; the "generated projections" sentence names exactly the files that moved; "the key the board already reads" checks against ListView.tsx:1464 / ObjectKanban.tsx:303 at the pin. Judged RIGHT.

③ Boundary flags

• (a) packages/lint gap — CONFIRMED on the head, and it is a metadata-authoring trap rather than a cosmetic note. validate-list-view-field-refs.ts:317-322 POSITIONS.kanban lists groupByField (error), summarizeField (warning), columns (warning) and no titleField, while calendar/gallery/map warn and gantt/timeline error on the same key. No test pins POSITIONS against the schema (the test table at validate-list-view-field-refs.test.ts:199-280 is hand-maintained), so CI stays green — and from this release a misspelled kanban.titleField is accepted by the spec, reported by nothing, and at the pin silently degrades to the resolver title (ObjectKanban.tsx:334-341). Demonstrated on the head lint sources with sibling controls: kanban: { titleField: "nope_field" } → no finding; the identical typo on kanban.summarizeField → warning, calendar.titleField → warning, gallery.titleField → warning, timeline.titleField → error (the lint suite itself: 112 passed). That is the ADR-0078 silently-inert shape; Prime Directive chore: version packages #10 says file, and the PR only noted it. Needs its own card (one-line fix: titleField: 'warning' in POSITIONS.kanban plus a test row). This seat's proxy refuses the search endpoint, so whether a card already exists is unverified — the adopting seat files or links one. Out of this card's declared scope; not a reason to refuse the diff.
• (b) objectui#7742 conditional resolves to "no" — verified: component.zod.ts:2846 ObjectKanbanPropsSchema.titleField already exists as an optional legacy alias of cardTitle on a different schema with a different semantic, so one declaration cannot serve both faces and the card is correctly not widened. Agreed.
• (c) KanbanConfigSchema carries no object-level .describe() where its siblings do — cosmetic, generated-docs only, out of scope. Agreed.
• (d) NEW — renderer rung ahead of the ruled fallback (① item 6): objectui pin ObjectView.tsx:2388 || 'name'. When the objectui leg of objectui#8367 declares the member and narrows the refused residual (KNOWN_REFUSED_RESIDUAL in ObjectView.kanbanGroupByRetired-8213.test.tsx:102 → ['cardFields']), it should also drop that floor as objectui#7029 did for calendar, so that "omit" reaches the ADR-0079 chain directly and the describe is exact. Sibling-repo work; escalated, not blocking here.
• (e) NEW, cosmetic — the new TSDoc docblock's chain (titleFormat → displayNameField → derivation → 'Untitled') is copied verbatim from Calendar's [spec: appearance.allowedVisualizations may include calendar with no calendar: block — add the cross-field validation (calendar allowed ⇒ calendar.startDateField required) #13817] docblock via the ruling's own quotation; ADR-0079 D2/D3/D6 in this repo and objectui record-title.ts:28-40 order it nameField first and floor at Record #id, with 'Untitled' only for an id-less record. Not in the .describe() that ships to the reference page; leave it with the sibling docblock it copies.
• (f) Downstream unblock is installability, not merge: objectui#8367 stays blocked until a released @objectstack/spec carrying this is pinned in objectui. Nothing for this PR to do.
• Gate families this seat could not measure here: check:react-declaration-parity (needs the objectui manifest — declared by check:generated as external input), the lint package's own vitest run (measured separately, see (a)). CI on the head: 34 success / 5 skipped / 0 failure, including Spec property liveness, Lint & Repo Gates, Test Core (6/6), TypeScript Type Check, Governed Surface Queue Guard.

Implemented-by: claude/issue-16894-kanban-titlefield
Reviewed-by: session_01JbZnqu8bt6YqfJsr9vaFb3

VERDICT: PASS

---

Generated by Claude Code

[os-bill]

Clause-② adopted — PASS; both carriers stripped in this act

Review of record: PR-thread comment 5710778226.
Head judged: 6d01b4b63e07b5fbdc0069ce8f5143167b9813ff — re-read live at adoption and unmoved (draft: true, mergeable_state: clean), so the head judged is the head that would merge.

Adopted verbatim. A dispatching seat may adopt an isolated reviewer's record whole or void it whole; it may ⛔ never rewrite or polish one, and this seat did neither. Everything below is disposition of the flags the record raises — ⛔ not a supplement to its judgments.

needs:contract-review is stripped from both carriers in this same act: PR #18561 and card #16894.

Independence pair

• Implemented-by: claude/issue-16894-kanban-titlefield — the branch that produced the diff (mode:subagent).
• Reviewed-by: session_01JbZnqu8bt6YqfJsr9vaFb3 — the seat that renders/adopts. An isolated review subagent holds no session of its own, so contract-review.md records the adopting seat.
• Different identities ⇒ no SELF-REVIEW.

Why an at-tier review exists here at all

This seat measures below CONTRACT_REVIEW_TIER (scripts/pm/dispatch-gates.mjs), so it is ⛔ barred from self-reviewing a clause-② card. It ran an isolated at-tier reviewer under the rule in #18511 — 「未达档席 ⛔ 不自审，起隔离达档子代理转录核档采信」 — on the maintainer's explicit instruction for this PR. That rule is not yet merged; this is stated plainly rather than implied.

Tier evidence, in the shape that rule demands: the reviewer parsed every type:"assistant" line of its own transcript for the harness-stamped per-request message.model and compared each against CONTRACT_REVIEW_TIER imported live from dispatch-gates.mjs — 30/30 at tier at posting, 32/32 on the final line, exactly one distinct model value, no fallback line anywhere in the run. ⛔ get_session was not used: inside mode:subagent it measures the dispatching session and is not mutual proof. The Served-tier: line of the record carries that 30/30 stamp control, and check-clause2-carriers --pair 18561 re-reads it at C7.

The reviewer was fed the card, the ruling it executes, and the PR body/diff — ⛔ not the dispatch order and ⛔ not this seat's own acceptance conclusions.

Disposition of the reviewer's boundary flags

• (a) POSITIONS.kanban carries no titleField row. The reviewer confirmed the gap first-hand on the head and asked the adopting seat to file or link, its proxy having refused the search endpoint. ⛔ No twin filed — the card already exists as [finding] POSITIONS.kanban is the one item-titled face with no titleField row — from #16894 the key is authorable and a misspelt one is walked by nothing, while the identical typo on calendar/timeline errors #18565, opened by this seat before the review and re-measured first-hand there (same sibling controls: the typo warns on calendar/gallery, errors on timeline, and reads as nothing on kanban). [finding] POSITIONS.kanban is the one item-titled face with no titleField row — from #16894 the key is authorable and a misspelt one is walked by nothing, while the identical typo on calendar/timeline errors #18565 is unlabelled for domain:* and ungraded; routing and grading are the triage seat's. Out of this card's declared scope, and ⛔ not a reason to hold this PR.
• (b) objectui#7742 resolves to "no" (component.zod.ts:2846 already spells an unrelated optional titleField on ObjectKanbanPropsSchema) and (c) the missing object-level .describe() is cosmetic — both agreed as the PR states them; nothing owed here.
• (d) NEW — renderer rung ahead of the ruled fallback. At the objectui pin, app-shell/src/views/ObjectView.tsx:2388 floors an omitted kanban.titleField at the literal 'name', so "omit ⇒ ADR-0079 chain" is exact only for records with no non-empty name. objectui-lane item, to be dropped on the objectui leg of objectui#8367 as objectui#7029 already did for calendar. Escalated, ⛔ not blocking a spec declaration.
• (e) NEW, cosmetic — the new TSDoc docblock's chain order is copied verbatim from Calendar's spec: appearance.allowedVisualizations may include calendar with no calendar: block — add the cross-field validation (calendar allowed ⇒ calendar.startDateField required) #13817 docblock via the ruling's own quotation, where ADR-0079 D2/D3/D6 order it nameField first and floor at Record #id. It does not reach the shipped .describe() or the reference page. objectui/sibling-docblock item; left with the docblock it copies.
• (f) Downstream unblock is installability, not merge: objectui#8367 stays blocked until a released @objectstack/spec carrying this key is pinned in objectui. Nothing for this PR to do.

Landing pre-checks, read on this head

#	check	reading
①	in-seat clause-② PASS on record	comment 5710778226, at-tier, Head-sha matches
②	double carrier	check-clause2-carriers.mjs --pair 18561 → exit 0 (both limbs agree, no widening tell, record on head) — re-run after the strip
③	PR checks	39 check-runs: 34 success · 5 skipped · 0 other; check-expected-skips.mjs --pr 18561 → exit 0, all 5 skips in the roster (Auto Label, Check PR Size, Console Pin Gate, Packed-tarball smoke (opt-in) ×2)
—	governed surface	check-governed-merges.mjs --pr 18561 → 0 of 7 paths hit the register ⇒ NOT governed; ordinary queue landing applies, ⛔ not the governed four-piece

Next: ready → auto-merge, then followed to MERGED.

[EDIT — added after posting, ⛔ not a re-measurement dressed as the original] The sentence above says the #18511 rule is "not yet merged". That was true when this comment was posted and is now stale: origin/main was re-fetched for unrelated work minutes later and its tip reads ab1d359259 skills(pm-dispatch): off-tier seat discharges clause-② review by spawning an at-tier subagent (ruling C) (#18511) — the rule is MERGED. Nothing about the review or this adoption changes; the rule this seat invoked simply stopped being provisional. Recorded rather than quietly edited away.

---

Generated by Claude Code