test(spec): the second root-level test file group's titles state each cited decision in words instead of a tracker number (stage 29)

[objectstack-fleet]

Part of #20749
Clause-②: no

Stage 29 of this card: the next area of class (e), the test strings shipped under packages/spec/src, as ruled in 5902360492 on #20513. This stage takes the second name-ordered group of the test files directly in packages/spec/src/: 15 files, root-entry-migrations-split.pin.test.ts through stack.test.ts. 12 of them carried 57 tracker ids in test titles (59 ids), one of which sits in a declared describe label rather than a literal title. All 57 now either state what their record decided, in words (form D), or drop the number where the title already says it. Text only: no assertion, identifier, test count, code comment, file name or non-test file changes. No file directly in src/ carries an id in a string any more (type-alias-convention.pin.test.ts, which sorts after the group, reads 0); the card stays open for the subdirectory files named under Acceptance notes.

Census (re-taken first)

The instrument is stage 28's census28.cjs, byte-identical (md5 31d8488b5194b8d3048e3fcaec0efaed, the value stage 28 published): an AST walk over the packages/spec/src test files, one message per folded string (a lone literal, a template, or a plus chain) that matches the gate's id pattern, a title when the folded root is argument 0 of a describe / it / test / suite / bench call, comments never read. It was run against the three published readings before it was trusted, and all three reproduce exactly: 128 messages / 130 ids in 37 files at stage 27's landing f7b8a5932b, 191 / 200 in 53 files (titles 162 / 168, other 29 / 32) at aa09db58c9, and 73 / 76 in 24 files at stage 28's landing b7e01fbbd.

reading	messages / ids	files
the group, base 43fc50051	57 / 59 (titles 56 / 58, other 1 / 1)	12 of 15
class (e) whole, base 43fc50051	73 / 76 (titles 61 / 64, other 12 / 12)	24
class (e) whole, this head	16 / 17 (titles 5 / 6, other 11 / 11)	12
the group, this head	0 / 0	0 of 15

The group reads 57 / 59, stage 28's landing figure: no re-cut. The census at 3f80f1716 (the dispatch base) and at stage 28's landing read the same per file as at 43fc50051.

Per file, messages at base: root-entry-migrations-split.pin 1, stack-artifact-crossref 11 (12 ids), stack-artifact-packages 3, stack-cross-reference-envelope 2 (one title, one declared label), stack-dev-logins 4, stack-email-template-locale-floor 6, stack-inline-action-crossref 13, stack-json-stage-package-body 4, stack-refusal-envelopes 2, stack-requires 4 (5 ids), stack-top-level-strict 5, stack 2. stack-conversions-record, stack-duplicate-action-key and stack-provenance carry no id in any literal.

Controls:

• Pathspec: the 15 named paths hit the control word describe( in 15 of 15 files, and a nonsense word in none; the census scanned 15 of 15.
• Lit outside the group, at base and head: ai/build-progress.test.ts 2 / 2 and ui/dashboard.test.ts 2 / 2.
• Lit at base and dark at head, inside the group: stack-artifact-crossref 11 / 12, stack-inline-action-crossref 13 / 13, stack-requires 4 / 5 and stack-cross-reference-envelope 2 / 2 each read 0.
• Dark at both ends: stack-conversions-record has no id anywhere; stack-duplicate-action-key and stack-provenance carry an id in one comment line each and read 0. The 15 group files read 0 at head while 55 of their comment lines still carry ids (112 id-bearing lines at base, comments included, 14 files).
• Planted, in a scratch tree: an id in a describe title, a plus-chain title, an expect message, an it.each row name, a template title and a declared string each read once (6 / 6); a comment, a six-digit number, an HTML entity and a markdown heading read 0.

What changed

57 literals, one line each, in 12 files: +57 / -57. Every file keeps its line count.

• 6 state the decision in words: "a modal target names a page, only" for the inline and the embedded modal-target refusals (stack-inline-action-crossref 133 and 330); "it was never built by a stack producer" for the hand-built stack object composeStacks refuses before either artifact-scoped rule runs (stack-artifact-crossref 380); "compatibility with the multi-package artifact" (stack-artifact-crossref 270); "unchanged from before the dangling-objectName refusal" (stack-inline-action-crossref 399); and the declared label "hooks[].object (the hook-ownership rule)", the rule ADR-0130 states as "the split must follow hook ownership" (stack-cross-reference-envelope 91).
• 4 get their subject back where the number was the only subject: "the no-floor report: warn-once bookkeeping" and "the floor guard's declared scope boundary" (stack-email-template-locale-floor 159 and 179), "the strict top-level door: the accept side does not move" and "defineStack surfaces the unknown-top-level-key refusal" (stack-top-level-strict 173 and 225).
• 47 drop a number the title already explains, with its connector: a #N — or #N prefix, a [#N] prefix, or a trailing (#N). That includes the twin title stack-artifact-packages.test.ts:355, which read does not warn about an undeclared composition rule (#5005 rule 3) and now reads does not warn about an undeclared composition rule, the same text stage 28 gave its copy at compose-stacks-manifest-preserve.test.ts:308.

Cited records

26 distinct ids. 25 answer 200 and were read: the body and the comment thread as the API serves it, with the rulings and landings that the rewritten titles state read in full. #10485 answers 404; its decision was read from its landing commit 35ad101bc (the themes carrier retired under ADR-0049 enforce-or-remove; app.branding is the one colour surface) and from the unknown-key strictness ledger, which records the maintainer's 2026-08-21 disposition as retiring the authorable surface, the phrase the title keeps verbatim. The comments counter of four records differs from what the comments endpoint serves (#14122 serves 21 of 22, #17614 5 of 6, #16449 0 of 5, #18056 5 against a counter of 4); everything served was read, and #16449's decision is in its body.

What each record decided, as the titles now say it:

• showcase's registered showcase_new_task action targets a PAGE, so the global_nav "New Task" command opens the home page in a dialog instead of creating a task #6739, ruling A (2026-08-09): a type: 'modal' target names a page, only; the object fallback is consumer leniency and enters retirement.
• [finding] os validate runs only the stack schema parse, so a config that exports a plain object (no defineStack() call) skips every defineStack cross-field refusal and passes #20367, ruling B: one authoring shape; defineStack and composeStacks stamp a provenance mark and composeStacks refuses an unmarked input with STACK_PROVENANCE_MISSING / 422.
• [finding] objectName on an object-embedded action is inert AND unvalidated — the third arm PR for #7397 deliberately left open #7456, Option A: the embedded objectName gets the registered rule's existence check, and only a dangling name newly refuses.
• packages: null on a release artifact: the schema and composeStacks refuse it, while every reader reads it as absent #19926, ruling A: packages: null is malformed, not absent.
• [finding] composeStacks drops a stack's grants and seed rows when its permissions / data is not an array — the composed artifact is silently incomplete, which ruling 5690859601 forbids (follow-up C of #18239) #19784, ruling C: a non-array concat key is refused with the envelope.
• composeStacks silently drops every non-array top-level key — api: today, server: as of #4910 #5005 (2026-08-04): same value passes, a conflict throws a prescriptive error, an undeclared top-level key warns.
• An unknown top-level stack key is named but still not rejected — and the diagnostic is not a warning, so objectstack validate --strict cannot catch it either (17.0.0 GA) #8687, Shape B: the top-level stack door is strict, and the lint yields to it.
• One artifact, N packages: let a release bundle carry co-owning packages so a product can be split into modules without renaming objects #14122 / spec: defineStack's per-stack cross-reference validation refuses permissions[].objects and data[].object that name an object owned by ANOTHER PACKAGE OF THE SAME ARTIFACT — so ADR-0130's co-owned artifact and the "permission sets stay whole in the app package" addendum cannot both hold once the app package owns objects #18202: one artifact carries N packages; the artifact-scoped permission and seed references resolve against the artifact, a stack that does not opt in is untouched.
• finding(spec): AssembledPackageBodySchema declares callable and custom branches an inert-JSON artifact cannot hold, and every schema embedding it loses its JSON Schema #17518, ruling A prime: the artifact-stage and record-stage package bodies are declared beside the assembled one.
• Inline (page-element) actions bypass defineStack's action cross-reference validation entirely — a dangling type: 'modal' target builds clean #6889 and [observation] config.objects[].actions[] (object-embedded actions) may also bypass defineStack's cross-reference walk — unverified, same mechanism as #6889's row E #7397: inline page-element actions and object-embedded actions get the registered action rule's target cross-references.
• spec: retire the page member of the list-view type enum and its pageName mount — enforce-or-remove, the objectui render half was never built (maintainer 2026-09-09 「撤」) #17063: the type: 'page' list-view mount is retired. Let an app contribute its own first-run credentials to the boot banner (devHint / devLogins[]) — the platform cannot know which of an app's audiences shows something, and only the app can #17556: an app declares its own devHint / devLogins. [finding] An emailTemplates bundle tagged with the stack's own defaultLocale (e.g. en) has no fallback floor: the resolver retries only the literal en-US, so every unlisted recipient locale dead-letters permanently #17614 and [finding] packages/spec gives two different rung counts for email-template locale resolution, and the floor guard has two silent early-returns that skip exactly the shapes it exists to catch #18056: the no-en-US-floor report and its declared scope. finding(spec): defineStack's cross-reference refusals are bare Errors — no ADR-0112 code / status — so five REFUSED item classes in the ADR-0130 matrix are distinguishable only by message text #14552 and finding(spec): six of defineStack's seven refusals are still bare Errors — no ADR-0112 code/status — leaving one envelope among neighbours that have none #15963: every defineStack refusal carries an ADR-0112 envelope; spec: register every error code that ships in dist — the nine unregistered defineStack / ownership / namespace codes enter ERROR_CODE_LEDGER / StandardErrorCode (the ledger is the published face, per the #16404 ruling) #16449 registers each shipped code in the ledger. Unify optional-plugin loading: framework serve.ts and cloud objectos-runtime are parallel implementations with divergent capability tokens (ai-studio vs aiStudio) #3265 and Remove deprecated aiStudio/aiSeat capability aliases after one deprecation cycle (follow-up to #3265) #3308: requires is validated at the producer and the camelCase aliases are removed. defineStack hard-errors on an undeclared hierarchy scope but is silent on an undeclared trigger capability — every autolaunched flow ships inert #14153 and [finding] requires: ['triggers'] without automation validates, but at boot the record-change trigger is not installed and every flow "will never run" #20332: trigger capability validation and its triggers without automation pair. spec: the root entry stops carrying the migration and conversion registries (about 1 MB of prose), so a console first screen no longer downloads os migrate text (the upstream payback of objectui#11088 D1) #20646: the root entry stops carrying the migration registries. defineMapping artifacts are registered but never consumed — wire named mappings into REST import or de-scope from the stack #2611: named import mappings are wired into the import path.

Readers

173 needles (each old literal, a 24-character window around each id, and the text left and right of each id) were searched with git grep -F over the tracked tree at the base outside the 15 files, with a lit control (composeStacks in stack.zod.ts, hit) and a dark control (a nonsense string, no hit). One needle hit, one line: a // comment in packages/qa/dogfood/test/fixtures/override-composite-fixture.ts that quotes the tail of one title with its id. It is a comment, not a reader. The only reader of the declared label is the file's own describe(row.label, ...). No group file reads a test name (currentTestName, expect.getState, snapshots: 0). The 15 file names appear elsewhere only in prose and comment lines, in packages/spec/test-typecheck-debt.json (keyed by file and error text, not by title) and in two dispatch-gates.mjs / check-changeset-no-major.mjs fixtures that name a path or a changeset file. Every old literal occurs exactly once across the 15 files (its own site).

Text-only proof

• textonly28.cjs (stage 28's, md5 957eff6b3837d762b8e03d070155930a), per file, base text against head text: both texts with every string token masked are byte-equal (code and comments), the string-token counts are equal, and every changed token sits in a describe / it / test title position or on a declared line, carried an id at base and carries no # plus digits at head. 15 / 15 files SAME; 57 changed tokens, 56 titles and 1 declared label (run with --declared 91); per-file counts exactly as predicted in writing at 2026-10-08T22:57Z, before the edit. The 3 untouched files read SAME with 0 changed.
• 17 / 17 controls on scratch copies of the head files behave as predicted: an identifier rename, a numeric literal, a comment edit, a non-title string given an id, a rewritten title given a new id, a title that was id-free at base edited, one rewritten title reverted to base (SAME, 0 changed), an it.each title given an id, an expect message changed, a title re-split into a plus chain, a template title given an id, a test added, an id appended to a rewritten title in a many-title file, an untouched file (SAME, 0 changed), the declared label run without --declared (refused), the declared line plus another changed string (refused) and the declared line alone (SAME, 2 changed).
• git diff -U0: 57 plus and 57 minus lines, each the planned line; 0 added lines carry # plus digits; control-byte scan of the 12 files and of this body: 0 hits.

Tests

• The 15 files with --project local --project repo and the JSON reporter, at the base (same worktree, before the edit) and at the head 65b42b7d1: 405 / 405 passed in 15 files, 108 suites each side; per-file test count and status sequence identical in 15 / 15; 229 full names changed, 0 mismatches against the plan (each head name equals the base name with the planned replacement applied); 0 duplicate full names either side; names with # plus digits 229 at base, 0 at head.
• spec vitest run --project local: 626 files passed, 18740 tests passed | 1 todo, exit 0.
• spec typecheck: VERDICT command-exit 0; check:test-typecheck 52 files / 246 errors / 135 pinned signatures held, unchanged; tsc --listFilesOnly -p tsconfig.test.json holds 15 of 15 group files.
• turbo build of all packages before the gates: Tasks: 71 successful, 71 total. Heavy runs went through scripts/pm/os-verify-lock.sh (slot dev-20749-s29).
• No reverse verification or ablation applies: no assertion, gate or reader moved.

Gates

All at 65b42b7d1. The dispatch list (77 commands, derived at 3f80f1716) equals the list re-derived from the real change set at this head, 12 paths against merge base 43fc50051 (114 changed lines): nothing added, nothing dropped. All 77 exit 0 and dispatch-gates --ran reads "77 derived, 77 run, 0 NOT-MEASURED, 0 UNRUN". Also run, exit 0: the five roster families that share a directory with the diff (check:meta-url-spelling, check:spec-changes, check:authz-resolver, check:error-code-casing, check:filter-alias-parity) and check:generated ("All 15 generated artifacts are up to date", measured against the dist the full build wrote at this head). ESLint, narrowed and proven: 15 files linted with --no-inline-config, 0 errors, 0 warnings; the population read from ESLint's own config is 15 configured, 0 ignored; no file has parserOptions.project or projectService, so type-aware linting is not enabled and this diff cannot move the verdict of a file it did not touch.

Packaging and landing

• No changeset: skip-changeset. npm pack --dry-run --json --ignore-scripts in packages/spec lists 2069 files, none of the 12 changed files and no *.test.ts; controls src/stack.zod.ts, dist/index.mjs and package.json are present; an old and a new title phrase read 0 files in dist/, the control Unrecognized key reads 42.
• Not governed: check-governed-merges --test on the 12 paths reads 0 of 12; 114 changed lines.
• git merge-tree onto origin/main 6a53564b9: clean. None of the 17 open PRs touches any of the 15 files (file lists read; 6 of them touch packages/spec/src, so the scan sees spec files).

Acceptance notes

• Class (e) is not finished: 16 messages / 17 ids remain in 12 files, all in subdirectories. By directory: ui/ 9 / 9 in 7 files, automation/ 2 / 3 in 1, ai/ 2 / 2 in 1, and api/, contracts/ and kernel/ 1 / 1 in 1 each. Their plan is the cutting seat's: stage 27's landing sets the needles stage apart (one stage, with an at-tier review) and keeps ui/'s four colour literals as decided, and this stage measured neither.
• Regrowth for stage 30's plan: five subdirectory test files carry 6 messages / 7 ids that were added after stage 27's landing. The files are older; the ids entered their strings in three commits. api/meta-item-response-shapes.test.ts 1 / 1 (a title, from feat(meta): the /meta item read serves the version token, If-None-Match: * pins a first write, and the 409 carries currentVersion #22126); automation/builtin-node-config.test.ts 2 / 3 (two titles, from feat(spec)!: flow value slots refuse the {…} template dialect, naming the CEL spelling of each token (#19939, C half) #22259, which re-added ids to a title stage 10 had rewritten); kernel/manifest.test.ts 1 / 1, ui/action-description.test.ts 1 / 1 (titles) and ui/view-submit-redirect-url.test.ts 1 / 1 (an other string, to be needle-checked there), all three from fix(spec): author-facing describes and refusals drop service-interface names, ruling dates and foreign example ids #22125. Nothing regrew since stage 28's landing: the per-file census at b7e01fbbd equals the one at 43fc50051. Test files are outside check:doc-authoring's ledgered leg, so nothing refused these on arrival. Not edited here.
• Code comments in the 15 files still carry live ids (55 lines). Comments are not this card's share. Carrier: none.
• origin/main moved from 43fc50051 to 6a53564b9 while this ran; one commit touches packages/spec (api/error-code-ledger.zod.ts, none of the 15 files), so nothing was merged into the branch.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

Nothing in this diff resolved to a documentable surface (no symbol, route or SDK anchor derived from 0 changed package(s)), so this run has no opinion about the docs.

What this run could not see

• 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.
• a key NAME is not a key, so the hand re-read the line above prescribes can land on the wrong schema. The same spelling is authorable on one governed type and a [REMOVED] tombstone on another for each of active, aria, joins, objects, template, tools and version (censused on [finding] tools is a key on BOTH AgentSchema (tombstoned, dead) and SkillSchema (live, cloud-attested), so a name-based search attributes skill examples to the agent key — it produced a false stop-the-line alarm on PR #19059 #19093 over the liveness ledger's governed types, top-level keys); nothing in a search result distinguishes the two, so a grep hit on a LIVE example reads as evidence about the DEAD key. Measured on fix(spec): the agent.tools liveness row says dead — it claimed live on a key the schema tombstoned #19059: content/docs/ai/agents.mdx was reported as contradicting the agent.tools tombstone over its tools: example at :161, which is inside the defineSkill({ block opened at :155 — the page was already correct. Settle ownership by PARSING the value against both schemas, never by the name: that literal PASSES SkillSchema, and as an AgentSchema it FAILS at tools with the tombstone prescription. ⛔ These names are not the whole class — a key retired through a .strict() guidance map leaves no tombstone in the walked shape and none of them here (tool.category, live as AIToolDefinition.category).

Coarse fallback — 0 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 345d3f3d86305eb3b5614a1788ef46598648e4f5 → packageMentionDocs.

[objectstack-fleet]

CI on 65b42b7d1: Test Core (3/6) red on a runtime test's 5 s timeout, not this PR's; one re-run by a merge-main push · domain:spec seat 2 (#18549) · session session_01DhTqaEHqPVSVnAkjG3jywn · 2026-10-09T00:35Z.

• Signature:
  • Failed step: "Run this shard's tests".
  • Error: @objectstack/runtime src/package-uninstall-org-scope.integration.test.ts:347 ("allows an org-scoped uninstall that spells allTenants:false explicitly"), Error: Test timed out in 5000ms. No assertion failed.
  • Counts: 1 failed / 4775 passed in the package; 14265 tests accounted for on the shard.
  • The shard's import phase took 1096.9 s, a loaded runner.
• The three facts for the one re-run, each read:
  1. The failing file's import closure is disjoint from this diff, which changes string literals in 12 packages/spec/src/*.test.ts files.
  2. main is green on Test Core (3/6) on its last 8 commits. The test file is unchanged since 2026-08-12.
  3. The first error is a timeout, not an assertion.
• Action: the stage's dev merges origin/main and pushes, which is the one re-run. A second red on the same signature is not re-run. It goes to the triage seat as this test's timing margin, the third load-timeout family member this week after hourly full run: red on main (CI) #22292 and hourly full run: red on main (CI) #22346.

---

Generated by Claude Code