[finding] field.zod.ts option-description docblock 与 field-rows-option-description.test.ts:25 断言「canonical 拼写 depends_on 已存在于字段级」——对 spec 自身为假(字段级现行声明键是 camelCase dependsOn)

Found during objectui#6140-半边收口(dev 报告 objectui#6140 comment 5481485473);filed unassigned 供分诊。

The shape

两处(#13669 落地物)散文断言「the canonical spelling depends_on already exists at the field level」:

packages/spec/src/data/field.zod.ts 的 option-description docblock
packages/spec/src/data/field-rows-option-description.test.ts:25(行号会 rot,按内容找)

实测为假:spec 字段级现行声明键是 camelCase dependsOn(field.zod.ts:1304,自 #2125;strict 门拒 snake;ledger LIVE)。「canonical depends_on」只对 option 面成立。该错句实际误导了一次派发(objectui#6140-半边任务 3,已 fork 到 objectui#6153 决策项)。

修复方向

两处一行改写(把「字段级」改为「option 面」或按 objectui#6153 裁定后的最终表述)。⚠️ 措辞终稿软阻塞于 objectui#6153 的 dependsOn 拼写裁定——但「现句为假」不依赖该裁定,可先改为仅陈述 option 面事实。

Refs

objectui#6153(决策项)· objectui#6140 comment 5481485473(证据链)· #13669(错句引入现场)

Claim: PM loop, domain:spec seat (seat post #6017), R3 of this shift
Session: session_0174WZTU6XcFcS7g2kykC53i (GitHub zhuangjianguo)
Branch: claude/issue-13916-option-dependson-docblock-fact
Worktree: objectstack-issue-13916
Domain: domain:spec
File surface: packages/spec/src/data/field.zod.ts — the option-description docblock sentence at :281-284 on a4e4d2d7 ("Per the same inherited ruling, dependsOn is NOT declared here — the canonical spelling depends_on already exists at the field level, and a camelCase twin would be a second spelling for a declared concept"), FACT LAYER ONLY: state what is true (the field-level key is camelCase dependsOn, :1327; depends_on is canonical on the option face only; whether the option face declares dependsOn is pending objectstack-ai/objectui#6153) and ⛔ invent no replacement justification for the not-declared conclusion · packages/spec/src/data/field-rows-option-description.test.ts:25 (file-header docblock, same sentence — a comment, not an assertion) and :205 (an inline comment "field-level depends_on); declaring a camelCase twin here is …", a third site triage did not list, same fix) · changeset: skip-changeset label on the PR for a comment-only change in shipped source, unless check:generated moves a published artifact, in which case a patch changeset for @objectstack/spec instead — the dev reports which. Stop on breach; explain in the report.
Container & model: S, mode:subagent, model: opus. node scripts/pm/dispatch-gates.mjs --tier on the surface at a4e4d2d7 (18:05Z): "no path-derived mandate" + "Clause ② SUSPECT surface" (both files under packages/spec/src/**); tier judged from content.
Clause-②: no — comment text only; no accept/reject or public-surface change.
Serial constraints cleared: open-PR file scan at 18:04Z (31 open PRs, 233 changed-file rows) — none touches field.zod.ts or the test file; same-day churn: YES — field.zod.ts last changed today at 14:28Z (f8e8f035): the dev works from merged main and locates the sentence by content; lane siblings in flight (#14847 stack.zod.ts, #14791 packages/lint, #14825 knowledge-source.zod.ts) and this batch's siblings (#13988 driver.zod.ts + three docs pages, #14048 filter.zod.ts + references/data/filter.mdx) file-disjoint; H17 on-hold trigger-file index (#9857 body, 13:47Z): #8345 declares field.zod.ts as a Restart-touch trigger — notification posted on #8345 in this stroke per its precedent, no rider taken (zero currency surface), its v18 Restart-when: legs re-read false at 18:14Z; no remote branch exists for this card (18:11Z).

Decision re-read (1 comment): triage 5486900363 — fact layer now, ruling layer (option-face dependsOn) ⛔ not in this card and owned by objectui#6153 (re-read 18:05Z: still open, pm:queue + domain:spec, no ruling yet — so the wording states only the option-face fact); the test-file site is a docblock, no test goes red; #13669 resolves to a PR, not an issue (merge commit 86e765a7 carries "(#13669)") — the dev corrects that provenance in the PR body. Premise re-read on origin/main a4e4d2d7 (18:05Z): the sentence sits at field.zod.ts:283 (moved from :272), :1327 declares dependsOn: at field level, the test file carries it at :25 and :205; depends_on occurs nowhere else under packages/spec/src — the three prose sites are its only occurrences. Premise holds.

Generated by Claude Code

Dispatch (R3 of this shift, 2026-09-03T18:17Z) — domain:spec seat, session_0174WZTU6XcFcS7g2kykC53i, seat post #6017. mode:subagent, model: opus, size S, Clause ② no (path limb fires, declared in the PR body). The dev leaves its own Claim: comment below; the assignee stays the PM's.

Dispatch text handed to the dev, in brief (triage 5486900363's two-layer split — fact layer now, ruling layer owned by objectui#6153 — and its "⛔ 实施者不得顺手编一个新理由把结论保住" are quoted verbatim and marked non-renegotiable):

Rewrite the false sentence at its three sites (field.zod.ts docblock; the test file's header docblock and its inline comment at the :205 region) to state only what is true: the field-level key is camelCase dependsOn; depends_on is canonical on the option face only; whether the option face declares dependsOn is pending objectui#6153. No replacement justification.
Mechanism hypotheses to verify: no generated artifact reads that JSDoc (check:generated clean without changes ⇒ skip-changeset; otherwise a patch changeset); no test asserts the sentence.
Provenance fix in the PR body: feat(spec): declare rows (multiline editors) and options[].description as authorable field-metadata keys (objectui#6140 / objectui#6153, spec half) #13669 is a PR (merge commit 86e765a7), not an issue.
Landing: draft PR Fixes #13916, two-limb Clause ② declaration, ⛔ no needs:contract-review, ⛔ no ready flip; report as os-dev-report on this card. H17: currency: no exchange-rate / home-currency conversion — dynamic multi-currency amounts cannot be reconciled or aggregated across currencies #8345 notified (5530136440), no rider taken.

Generated by Claude Code

os-dev-report

{
  "issue": 13916,
  "status": "done",
  "branch": "claude/issue-13916-option-dependson-docblock-fact",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/15055",
  "premise_still_valid": true,
  "summary": "Rewrote the false sentence at all three prose sites (the option-description docblock in field.zod.ts, and the pin file's header docblock plus its inline comment) to state only measured facts, with no replacement justification for the not-declared conclusion; no schema member and no expect() moved, so the option shape accepts and refuses exactly what it did before. Measurement refuted one dispatch instruction: it asked the new wording to say depends_on is the canonical spelling on the OPTION face only, but the built schema refuses depends_on on the option face too, and on the field face with a rename hint to dependsOn - depends_on is declared nowhere in @objectstack/spec; it is objectui's BaseFieldMetadata spelling (packages/types/src/field-types.ts:128), so the landed prose says that instead. H1 holds (check:generated clean without --fix once dist was rebuilt), so no changeset: skip-changeset was applied to the PR and confirmed present on a later read alongside three auto-labels. Dedupe surfaced #14583, an unassigned, unbranched duplicate card for the same sentence whose acceptance criterion (zero depends_on lines in field.zod.ts) contradicts this card's fence, and whose body records a maintainer ruling dated 2026-09-02 on objectui#6153 that post-dates triage 5486900363's open-and-unruled reading - both surfaced in the PR body and left to triage, not acted on.",
  "tests": "All at 9300bca5 (git status --porcelain and git diff HEAD both empty, so the pre-commit runs measured a byte-identical tree). Build: pnpm --filter '@objectstack/spec^...' build then pnpm --filter @objectstack/spec build - os-verify-lock VERDICT command-exit 0. Tests: pnpm --filter @objectstack/spec exec vitest run --maxWorkers=2 - 'Test Files 463 passed (463) / Tests 12374 passed (12374)'. Typecheck: pnpm --filter @objectstack/spec run typecheck - clean, incl. check:test-typecheck 'OK - @objectstack/spec's test layer compiles under packages/spec/tsconfig.test.json' (so the edited test file sat inside a program that was really checked). H1: pnpm --filter @objectstack/spec run check:generated (no --fix) - 'All 15 generated artifacts are up to date'; the first post-edit run reported api-surface stale only via its PREREQUISITE line 'dist/**/*.d.ts is OLDER than packages/spec/src', which cleared after rebuilding - NOT MEASURED, not a moved artifact. H2: git grep for 'already exists at the field level' and 'camelCase twin' across packages and scripts exits 1 (no match), positive control - a grep for a phrase the new docblock does carry returns its line. H3: post-change git grep -n depends_on -- packages/spec/src returns the 8 rewritten prose lines plus ui/view.test.ts:234 and :434, which are dependenciesField: 'depends_on' - a gantt fixture field MACHINE NAME (snake_case is correct for a data value), not a metadata key; those two are a delta against the dispatch's 'three prose sites are the only occurrences' reading. Gates: node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack re-derived on the real diff (its stderr warns the tree is 7 commits behind origin/main, 4 derivation inputs changed in that range - workflow/PM scripts, none of them a gate reading spec prose), then 32 gates run, all exit 0: spec check:docs, check:api-surface, check:authorable-surface, check:export-origins, check:liveness, check:llms-txt, check:skill-refs, check:strictness-ledger, check:variant-docs, check:yaml-examples, check:empty-state, check:exported-any, check:dual-source-exports, check:entry-nameability, check:browser-reachable-entries; repo check:nul-bytes, check:merge-driver, check:published-files, check:doc-authoring, check:cross-package-test-inputs, check:test-source-alias, check:spec-parsed-alias, check:partof-closing-keyword, check:changeset-gate-self-tests, check:objectql-double-limit, check:where-matcher, check:page-declaration-shape, check:slot-lookup, check:type-source-resolution, check:logger-receiver-detach, check:engine-double-contract; plus @objectstack/lint check:doc-formula-expressions, which first exited 3 on its own PREREQUISITE NOT MET path (nothing measured) and went green after building @objectstack/formula and @objectstack/lint. Every exit code was captured before any pipe. Name correction: the dispatch's check:closing-keyword-parity is spelled check:partof-closing-keyword here. Declared narrowing: the repo-wide pnpm lint sweep and the remaining families of the 204 discovered were left to CI. No ablation was run - the change is comment text with no behaviour to ablate.",
  "mcp_calls": "0 - every GitHub read and write used a zero-quota channel: the public issue page's embedded JSON payload for #13916's body, and the authenticated repo-scoped REST endpoints (probe green: GET issue 200) for comments, the open-issue dedupe listing (533 rows, 507 non-PR), objectui#6153, PR create, PR body patch, the additive label POST, and the finding issue",
  "open_questions": [
    {
      "question": "#14583 is an unassigned, unbranched duplicate of this card for the same sentence, and its acceptance criterion contradicts this card's fence: it requires 'grep -n depends_on packages/spec/src/data/field.zod.ts' to return ZERO lines, while this card's dispatch says the word may stay wherever it is true. Which end state governs, and who closes the loser?",
      "options": [
        "A treat #14583 as a duplicate of this card, closed by the PM, and keep this PR's wording, which uses depends_on only in measured true statements (where it is declared: objectui; where it is refused: here)",
        "B treat #14583's zero-occurrence criterion as governing and re-word this PR to name only dependsOn, dropping the reader's only pointer to where the snake spelling actually lives",
        "C land this PR and re-triage #14583 as a separate p3 for the remaining wording preference"
      ],
      "recommendation": "A, because the zero-occurrence criterion was written under the pre-triage reading in which depends_on had no true home in the prose; after the rewrite each occurrence is a measured statement, and deleting them would remove the one sentence that stops the next author from re-deriving the false claim. This is a triage call, not mine - I did not touch #14583."
    },
    {
      "question": "#14583's body records a maintainer ruling dated 2026-09-02 (decision batch on objectstack-ai/objectui#6153, recommendation A, verbatim agreement) that the field-level cascade key is dependsOn and that objectui's snake twin retires under enforce-or-remove. That post-dates triage 5486900363 and the dispatch's own 18:05Z re-read of objectui#6153 as still open and unruled. The landed prose says the option-face question is 'pending objectui#6153' - is that now stale?",
      "options": [
        "A leave the wording as landed (every clause is independently measured; only the pointer's tense is at issue) and let the ruling layer be recorded on objectui#6153 itself",
        "B amend the prose to cite the 2026-09-02 ruling on the strength of the second-hand record in #14583's body",
        "C ask the maintainer to confirm the ruling on objectui#6153, then amend in a follow-up"
      ],
      "recommendation": "A now, C next - writing a ruling into shipped source from an issue body that reports a live-chat decision is exactly the substitution triage 5486900363 forbade, and the dispatch's own re-read today disagrees with that record. Nothing in the landed prose is false under either reading."
    }
  ],
  "out_of_scope_findings": [
    "filed as #15058: the pending .changeset/field-rows-and-option-description-declared.md carries a fourth copy of the same false claim ('the canonical field-level depends_on already exists') and is release-notes input; not ridden here because editing .changeset/** pulls in gate families (check:objectui-changeset, check:pm-half-states) outside this card's verification surface, failing condition 4 of the bounded in-place exemption",
    "not filed, reported instead: #14583 duplicates this card (unassigned, no branch) and its body records the 2026-09-02 objectui#6153 ruling - closing or re-triaging it is the PM's call, and both open questions above turn on it"
  ]
}

Generated by Claude Code

Review — REWORK (one patch round) — PR #15055 (head 9300bca5, Fixes #13916), Clause ② not applicable (no on both carriers, check-clause2-carriers.mjs --pair 15055 exit 0 at 19:18Z)

Reviewer of record: PM seat domain:spec (session_0174WZTU6XcFcS7g2kykC53i). The dev's run is a good one: it refuted the dispatch's own instruction by measurement (depends_on is refused on BOTH faces of this package, with a rename hint on the field face — so "canonical on the option face" was my error, not the card's), corrected the PR #13669 provenance, kept every expect() untouched, found a fourth copy of the sentence in a pending changeset (#15058), and found the duplicate card. What sends it back is not a defect in the run but a ruling that neither the triage of 2026-09-01 nor this seat's 18:05Z re-read had seen.

The ruling record (read at 19:18Z). objectstack-ai/objectui#6153, comment 5510108593 by the director seat, 2026-09-02: maintainer ruling A, verbatim 「同意」 — the contract's field-level cascade key is dependsOn; objectui's BaseFieldMetadata mirrors it; objectui's snake_case depends_on twin is retired on its own card under enforce-or-remove. That amends the 2026-09-01 ruling whose premise was exactly the false sentence this card rewrites. The director filed the spec-side follow-up as #14583 (same sentence, same file), with the acceptance "zero depends_on lines in field.zod.ts" — a first-hand record, not a second-hand one.

Answers to the two open questions.

spec field.zod.ts option-shape comment claims a field-level depends_on the spec does not declare — correct the comment (objectui#6153 ruling A follow-up) #14583 is not a competing card but this fix's other name: folded into this dispatch as a family member (claim posted on spec field.zod.ts option-shape comment claims a field-level depends_on the spec does not declare — correct the comment (objectui#6153 ruling A follow-up) #14583 in this stroke; five gates hold). Its acceptance governs the wording, so option A-with-the-director's-criterion: the PR keeps the true statements and drops the snake spelling from field.zod.ts entirely.
"Pending objectui#6153" is stale under the ruling: the field-level spelling is settled (A), and the option face's state is a fact to state, not a question to defer — it declares description only and no cascade key of its own, per the inherited [skill] pm-dispatch: record two meta-judgments approved 2026-08-07 (silently-dropped declarations; two implementations of one operation) #6140 / fix(service-automation): runAs:'system' 的 create_record 按 ADR-0118 染全三列——组织、属主、创建者禁 NULL (#5494) #6153 ruling. Nothing in the landed prose is false today, but two of its sentences describe objectui's snake twin as a living declared spelling that the ruling retires; a shipped source comment should not point at a transitional state.

Patch round (sent to the dev directly; recorded here).

Reword the three sites to: the option shape declares description only and no cascade key of its own (the pin asserts the dependsOn refusal); the field-level cascade key this package declares is the camelCase dependsOn (the FieldSchema member below); objectui mirrors that spelling per the 2026-09-02 ruling on objectui#6153. ⛔ No depends_on token anywhere in field.zod.ts (the spec field.zod.ts option-shape comment claims a field-level depends_on the spec does not declare — correct the comment (objectui#6153 ruling A follow-up) #14583 acceptance); in the test file at most one factual mention that the snake spelling is refused with a rename hint. ⛔ Still no justification sentence for the not-declared conclusion beyond citing the inherited ruling.
PR body: a second line Fixes #14583 directly under Fixes #13916 (one closing line per family member; nothing else near a card number), the ruling citation, the updated before/after and post-change grep.
Re-run the union at the new head and quote the new short HEAD in the report and PR body; post a fresh os-dev-report here.

Everything else stands as reviewed: not governed (0 of 2 paths), trial merge clean, skip-changeset correct (check:generated clean without changes after a rebuild — no published artifact reads the JSDoc), 463 files / 12374 tests, typecheck clean, 32 gates green, H1–H3 confirmed with controls, MCP calls 0. The view.test.ts:234 / :434 dependenciesField: 'depends_on' hits are a fixture field's machine name, not a metadata key — the dev's classification is right and the dispatch's "only three occurrences" reading was the imprecise one.

Generated by Claude Code

os-dev-report

{
  "issue": 13916,
  "status": "done",
  "branch": "claude/issue-13916-option-dependson-docblock-fact",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/15055",
  "premise_still_valid": true,
  "summary": "Patch round on the same branch, head now daf2fd30. Re-worded all three prose sites to the 2026-09-02 ruling (verified first-hand at objectui#6153 comment 5510108593, recommendation A, maintainer verbatim agreement): the option shape declares description and no cascade key of its own per the inherited 6140/6153 ruling, the cascade key is FieldSchema's camelCase dependsOn, and objectui mirrors that spelling. Every 'pending objectui#6153' phrasing is gone and #14583's acceptance is met - grep -n depends_on packages/spec/src/data/field.zod.ts returns zero lines (instrument control: dependsOn returns 4), with exactly one snake_case mention left in the pin file naming the spelling FieldSchema refuses with a rename hint. Also folded in the CI red the previous head produced: the docblock reflow rotted the system-context census anchor, repaired by check-system-context-census.mjs --fix (field.zod.ts:1537 to :1540, one anchor, no refusal), so content/docs/permissions/system-context.mdx is now part of the file surface; origin/main was merged in (no rebase, merge committed before any gen:schema ran) and the census re-run green after it. PR body now carries the two closing lines, for #13916 and for #14583, as its first two lines, the ruling citation, all three before/after pairs, the zero-hit grep and the anchor evidence; skip-changeset stays and is confirmed on read-back. No expect() and no schema member changed.",
  "tests": "All at daf2fd30, run after the final commit and after the origin/main merge; git status --porcelain empty at that head. Build: pnpm --filter '@objectstack/spec^...' build then pnpm --filter @objectstack/spec build - os-verify-lock VERDICT command-exit 0. Tests: pnpm --filter @objectstack/spec exec vitest run --maxWorkers=2 - 'Test Files 465 passed (465) / Tests 12396 passed (12396)' (two more files than the pre-merge run, from the merged main). Typecheck: pnpm --filter @objectstack/spec run typecheck - clean, incl. check:test-typecheck 'OK - @objectstack/spec's test layer compiles under packages/spec/tsconfig.test.json'. check:generated (no --fix): 'All 15 generated artifacts are up to date', and the build's gen:schema left no tracked artifact dirty (git status empty afterwards). Anchor leg, all three as CI composes the step: pnpm check:doc-anchors ('298 internal fragment links across 409 source files all resolve'), check-system-context-census.mjs --self-test ('all cases passed'), and the census ('OK - 106 elevation read sites in 20 packages across 45 files, all anchored; 140 anchors resolve, 27 declared non-read'); the --fix run printed 're-anchored content/docs/permissions/system-context.mdx:199 field.zod.ts:1537 -> field.zod.ts:1540, 1 anchor(s) rewritten' and did not refuse, so nothing was hand-edited. Acceptance: grep -n depends_on packages/spec/src/data/field.zod.ts exits 1 with no output (control grep -c dependsOn = 4); git grep -n depends_on -- packages/spec/src returns 3 lines, none in field.zod.ts - the one permitted pin-file mention plus the two view.test.ts gantt fixtures where it is a field MACHINE NAME, not a metadata key. Gates: dispatch-gates.mjs --repo objectstack-ai/objectstack re-derived on the post-merge diff (the docs page pulls in the docs family), then 46 gate invocations all exit 0, exit codes captured before any pipe - the 32 from the previous round plus check:doc-anchors, check:skill-examples, check:corpus-claim-drift, check:docs-audit-scope, check:docs-redirects, check:docs-single-h1, check:published-readme-links, check:react-page-adapter-contract, check:role-word, check:skill-identifier-liveness, check:vendor-version-stamps, check:doc-security-posture, and the census pair. Three gates first exited on their own PREREQUISITE paths (NOT MEASURED, not findings) and went green after the builds they name: check:doc-formula-expressions needed @objectstack/formula and @objectstack/lint built; check:skill-examples needed @objectstack/client-react built with its dependency closure. Declared narrowing: repo-wide pnpm lint and the rest of the 204 discovered families are left to CI. No ablation - the change is comment text and one line anchor, with no behaviour to ablate.",
  "mcp_calls": "0 for the whole run, both rounds - reads via the public issue page's embedded JSON payload and authenticated repo-scoped REST (including objectui#6153 comment 5510108593, read first-hand before the re-wording), writes via REST: PR create, two body patches, the additive label POST, the finding issue, and these report comments",
  "open_questions": [],
  "out_of_scope_findings": [
    "filed as #15058 (still open, unassigned): the pending .changeset/field-rows-and-option-description-declared.md carries a fourth copy of the same false sentence and is release-notes input; still out of this PR because editing .changeset/** pulls in gate families (check:objectui-changeset, check:pm-half-states) outside this card's verification surface. Its body now needs a one-line update too - it cites the pre-ruling framing - which the taker can fold in.",
    "resolved this round, no longer an open question: #14583 is a family member of this dispatch and its acceptance governed the wording; it is named on the PR's second closing line, and the 2026-09-02 ruling replaced the 'pending objectui#6153' phrasing everywhere"
  ]
}

Generated by Claude Code

Review, round 2 — ACCEPT — PR #15055 (head daf2fd30, Fixes #13916 + Fixes #14583), Clause ② not applicable (no on both carriers; check-clause2-carriers.mjs --pair 15055 exit 0 at 19:59Z)

Reviewer of record: PM seat domain:spec (session_0174WZTU6XcFcS7g2kykC53i). Readings against GitHub, origin/main 89eb997d and the PR branch; UTC times are read times.

The REWORK's asks, verified on the branch (19:59Z).

Wording at all three sites states the 2026-09-02 ruling A (objectui#6153 comment 5510108593, read first-hand by the dev): the option shape declares description and no cascade key of its own, pinned by the refusal test; the cascade key is FieldSchema's camelCase dependsOn; objectui mirrors that spelling. No "pending objectui#6153" phrasing remains (PR body scan: absent). No justification sentence beyond citing the inherited ruling.
spec field.zod.ts option-shape comment claims a field-level depends_on the spec does not declare — correct the comment (objectui#6153 ruling A follow-up) #14583's acceptance holds: depends_on occurs 0 times in packages/spec/src/data/field.zod.ts on daf2fd30 (instrument control: dependsOn 4). Under packages/spec/src the token survives at exactly one prose site (field-rows-option-description.test.ts:208, the permitted "refuses the snake_case depends_on" mention) plus the two view.test.ts gantt fixtures where it is a field machine name.
PR body opens with Fixes #13916 / Fixes #14583 — one closing line per family member, no other closing keyword (the other numbers sit next to no verb). Provenance corrected to PR feat(spec): declare rows (multiline editors) and options[].description as authorable field-metadata keys (objectui#6140 / objectui#6153, spec half) #13669.
The CI red from round 1 is repaired the way the gate prescribes: check-system-context-census.mjs --fix re-anchored content/docs/permissions/system-context.mdx:199 from field.zod.ts:1537 to :1540 (one rewrite, no refusal, no hand edit), consistent with the docblock's net +3 lines; the census trio (check:doc-anchors, self-test, census: 106 sites / 140 anchors) green after merging current main.

Form, scope, evidence. Draft, base main; three files (field.zod.ts docblock, the pin file's two prose sites, the permissions page anchor); trial merge clean; check-governed-merges.mjs --test on the three paths: NOT governed (0 of 3); no expect() or schema member changed (SelectOptionSchema accepts and refuses exactly what it did). skip-changeset stands: check:generated clean without changes after a rebuild, the anchored page is not a published package artifact. Pinned to the head: 465 files / 12396 tests, typecheck clean with the test-layer ratchet, 46 gate invocations exit 0 (the docs family joined the derivation once the page entered the diff), prerequisite gates re-run green after building their packages. MCP calls: 0 across both rounds.

Out-of-scope group (sweep criterion: other copies of the falsified sentence). #15058 (dev-filed): the pending field-rows-and-option-description-declared changeset carries a fourth copy and is release-notes input — its body still frames the fix pre-ruling; the taker updates it in the same stroke. Left for triage.

Landing. CI on daf2fd30: 17 in progress, 0 red (19:59Z). At all-checks-green: provenance comment on the PR, flip to ready, auto-merge (squash), enqueue reading. On MERGED: both cards close via Fixes, strip pm:dispatched on each, probe the docblock and the page anchor on origin/main, landing note here with a pointer on #14583. Dev session: Agent task ae8a3129e974db63a (subagent, two rounds, finished; worktree removed without force).

Generated by Claude Code

Landed (PM seat domain:spec, session_0174WZTU6XcFcS7g2kykC53i, 2026-09-03T20:53Z) — PR #15055 merged as d261cefa at 20:51:42Z (squash via the merge queue; enqueued 20:25:35Z). This card and #14583 closed via the two Fixes lines (both closed_at 20:51:43–44Z, completed); pm:dispatched stripped on both in this stroke (read back: documentation / domain:spec / priority:p2 remain here).

Probed on origin/main d261cefa (20:52Z):

git grep -c depends_on -- packages/spec/src/data/field.zod.ts → no match (spec field.zod.ts option-shape comment claims a field-level depends_on the spec does not declare — correct the comment (objectui#6153 ruling A follow-up) #14583's acceptance criterion: zero lines); positive control dependsOn → 4.
The rewritten docblock is in place: "The cascade key this package declares is the camelCase dependsOn …" at field.zod.ts:284, citing the maintainer ruling 2026-09-02 on objectui#6153.
content/docs/permissions/system-context.mdx:199 cites field.zod.ts:1540 — the census-rewritten anchor, green in Lint & Repo Gates on the merged head.

Still open elsewhere, from the PR body: #15058 (the pending changeset .changeset/field-rows-and-option-description-declared.md carries a fourth copy of the falsified sentence — filed unassigned, release-notes input, not a rider here). objectui#6153's objectui half (the BaseFieldMetadata mirror and the snake_case twin's retirement) is the objectui domain:spec lane's dispatch, unchanged by this landing.

Generated by Claude Code

Navigation Menu

Repository navigation

[finding] field.zod.ts option-description docblock 与 field-rows-option-description.test.ts:25 断言「canonical 拼写 depends_on 已存在于字段级」——对 spec 自身为假(字段级现行声明键是 camelCase dependsOn) #13916

The shape

修复方向

Refs

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[finding] field.zod.ts option-description docblock 与 field-rows-option-description.test.ts:25 断言「canonical 拼写 depends_on 已存在于字段级」——对 spec 自身为假(字段级现行声明键是 camelCase dependsOn) #13916

Description

The shape

修复方向

Refs

Activity

zhuangjianguo commented on Sep 3, 2026

zhuangjianguo commented on Sep 3, 2026

claude commented on Sep 3, 2026

zhuangjianguo commented on Sep 3, 2026

claude commented on Sep 3, 2026

zhuangjianguo commented on Sep 3, 2026

zhuangjianguo commented on Sep 3, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions