spec: three source comments still describe the per-entry, self-contained bundle layout that #25748 replaced with shared chunks

Premise checked before dispatch (domain:spec seat 2, 2026-10-26): readings and corrections in comment 6409955774 (#25811 (comment)). The goal below is unchanged.

Filing gate ③, a follow-up owed by a landed change: PR #25748 (merged 3a2d38c6e2, card #8133), open question 1, option A. The seat's ACCEPT (6393800730, item 6) committed to this card at the merge. Filed by domain:spec seat 2 (seat post #18549) · sunhaolin · session session_01PC4NPXPgPmpDRG9DdDnyXT. ⛔ Not a claim. Grade and routing are triage's. The seat suggests domain:spec, p3, comment-only.

What is stale (measured on `main` `3a2d38c6e2`)

#25748 made @objectstack/spec's JS build one module graph with shared chunks. It kept packages/spec/src/** out of scope, so these comments still describe the old layout. Each is now wrong about the layout, not about behaviour:

packages/spec/src/index.ts:73: "Every subpath entry is a self-contained bundle, so a vocabulary symbol that …"
packages/spec/src/conversions/registry.ts:18242: "… entry is one flat file (tsup.config.ts, splitting: false), so a …". The PURE-annotation reasoning it supports still holds per chunk.
packages/spec/src/api/package-api-assembled.zod.ts:40: "… because the entry ships as one self-contained bundle and little of that tree …". This is TSDoc, so it renders into the generated content/docs/references/api/package-api-assembled.mdx.

What would change

Reword the three comments to the shared-chunk layout: a module two entries reach ships once, in a shared chunk; what an entry's graph reaches is still what a consumer may pay for. packages/spec/README.md and tsup.config.ts's oneModuleGraph docblock already say this.
Regenerate with gen:docs for the third.
No code, key or behaviour changes. A path under packages/spec/src/** is touched, so the contract-review tier applies to the docs face.

Dedupe

Searched repo:objectstack-ai/objectstack is:open self-contained bundle comment spec shared chunks: no open card covers it. New.

Generated by Claude Code

Triage: first grade — `documentation` · `priority:p3` · `domain:spec` · `area:devpath` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-25T15:57Z

⛔ Not a claim, ⛔ not a dispatch.

p3, as the seat suggested: comment-only, a follow-up owed by build(spec): one module graph with shared chunks, so each schema ships once and is one object across entries #25748. Regenerate gen:docs for the TSDoc one.

Generated by Claude Code

Premise check before dispatch · domain:spec seat 2 · 2026-10-26T06:43Z

Read-only premise reading on origin/main d5f4c2a2a (dispatch-runbook 派发前置). The goal of this card is unchanged.

Holds, one line number corrected: the stale comment in packages/spec/src/conversions/registry.ts sits at 18242-18243 ("Each published entry is one flat file (tsup.config.ts, splitting: false)…"). Sites 1 (packages/spec/src/index.ts:73) and 3 (packages/spec/src/api/package-api-assembled.zod.ts:40) are as the card says. The generated page content/docs/references/api/package-api-assembled.mdx:31 carries the same sentence and is regenerated with gen:docs, not hand-edited. Correct wording already exists at packages/spec/README.md:23 and packages/spec/tsup.config.ts:236-254.

Claim: PM loop round 1 · 2026-10-26T06:50Z
Session: session_01RyMY3prh9GRva1Tt37Di3i
Account: zhuangjianguo
Branch: claude/project-thread-mybepv (platform exception, written as measured: this seat session can push only its assigned branch, so the in-seat dev builds on it instead of claude/issue-25811-<slug>)
Worktree: objectstack-issue-25811
Domain: domain:spec
Seat: domain:spec#2
File surface: packages/spec/src/index.ts, packages/spec/src/conversions/registry.ts (comment at 18242-18243 only), packages/spec/src/api/package-api-assembled.zod.ts, generated content/docs/references/api/package-api-assembled.mdx via gen:docs (stop on breach; explain in the report)
Container & model: S 级机械卡, mode:subagent, model: TIER_FLOOR (sonnet) — --tier: no path-derived mandate; comment-only edit, effort: medium
Clause-②: no
Metadata-protocol: none
Responsibility: n/a — not a defect card
Thread-read: 6409955774
Premise: corrected 6409955774
Serial constraints cleared: #25822 and #25767 (dispatched this round) also edit packages/spec/src/conversions/registry.ts in other hunks; this comment-only edit merges trivially beside them. No other in-flight claim on these files.

os-dev-report

{
  "issue": 25811,
  "status": "done",
  "branch": "claude/project-thread-mybepv",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/25975",
  "head_sha": "1a17dab74385301324e630336650ddcf888d429f",
  "session": "session_01RyMY3prh9GRva1Tt37Di3i",
  "premise_still_valid": true,
  "summary": "Reworded the three stale per-entry-bundle comments (packages/spec/src/index.ts, the conversions/registry.ts PURE-annotation note only, packages/spec/src/api/package-api-assembled.zod.ts) to the shared-chunk layout of README.md and the tsup.config.ts oneModuleGraph docblock, and regenerated content/docs/references/api/package-api-assembled.mdx with gen:docs (one tracked file changed). Comment-only, no code, key or behaviour change; landing point equals the claimed file surface. Changeset (patch, @objectstack/spec) included: the new sentences are present in dist/index.d.ts and dist/api-assembled/index.d.ts, so the shipped files[] content moved and skip-changeset does not apply (measured criterion beats the fast-lane word comments).",
  "files_changed": [
    "packages/spec/src/index.ts",
    "packages/spec/src/conversions/registry.ts",
    "packages/spec/src/api/package-api-assembled.zod.ts",
    "content/docs/references/api/package-api-assembled.mdx",
    ".changeset/25811-spec-shared-chunk-comments.md"
  ],
  "line_budget": "none set; 37 changed lines (+26/-11), under the 3000 human-merge threshold",
  "tests": "No behaviour changed, no test added. pnpm --filter @objectstack/spec build via os-verify-lock: VERDICT command-exit 0. pnpm --filter @objectstack/spec gen:docs: exit 0, 'Generated 229 files', git status shows exactly content/docs/references/api/package-api-assembled.mdx modified. Built dist grep: new sentences present in dist/index.d.ts and dist/api-assembled/index.d.ts. No ablation (comment-only).",
  "gates": "Local: dispatch-gates --commands derives 108 families at head 1a17dab74; 105 run with exit codes captured per command before any pipe, all exit 0 (incl. check:docs, check:generated, check:api-surface, check:authorable-surface, check:upgrade-guide, check:issue-citations, check:doc-authoring, check:nul-bytes, check:changeset-no-major, check:empty-changeset; four prerequisite-gated ones exit 3 first, then exit 0 after building the lint and client-react closures). --ran reconciliation: 108 derived, 105 run, 3 UNRUN, all NOT MEASURED: pnpm check:dual-build-cjs-loads and pnpm check:lean-entry-closure (exit 3 PREREQUISITE NOT MET: they read built dist of packages this card does not touch; CI runs them), node scripts/check-plugin-teardown-shape.mjs --self-test (exit 3: shallow clone cannot read its pinned positive control; checker-health only). CI check-runs on the PR at report time are queued/in_progress (Check Changeset, Lint scoped lanes, Type Check lanes, Governed Surface Queue Guard, Check PR Size and others); only 'No other open PR may claim the same issue' and 'No other open PR may claim the same single-writer path'-class early jobs had reported, so every family CI runs is in_progress, not re-run locally.",
  "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm",
  "mcp_calls": "0 — no GitHub MCP calls",
  "api_writes": "2 — fleet-write relay dispatch of pr_create (creates the draft PR #25975 with assignees zhuangjianguo; one earlier dispatch of the same action was REFUSED by write-pace before writing anything, nothing stored) and POST of this os-dev-report comment via scripts/pm/post-stamped.mjs on issue 25811. Plus one git push (3 commits) to claude/project-thread-mybepv (not a REST write); the first push attempt was refused locally by the pre-push hook before reaching the remote.",
  "deviations": [
    "Order says commit trailers end with Co-Authored-By naming Claude Opus 5.5 (1M context); os-dev.md and the pre-push hook (check:commit-card-trailers) require the model-free pair, and the push was refused with the model-named trailer. os-dev.md wins per the order: commits carry Claude-Session plus 'Co-authored-by: Claude <noreply@anthropic.com>'. No model identifier is in pushed content.",
    "Branch is the platform-assigned claude/project-thread-mybepv (as the Claim comment states), not claude/issue-25811-slug; it already existed locally at an ancestor of origin/main and not on the remote, so no empty probe was pushed and the first push carried real work.",
    "Mis-sequenced command: a first 'cd ../objectstack-issue-25811' failed (worktree not yet created) and a chained pnpm install --frozen-lockfile ran in the seat checkout /home/claude/objectstack once. It touched only gitignored node_modules; HEAD (detached d5f4c2a2a) and git status stayed clean, no file edited there.",
    "PR carries no labels: skip-changeset deliberately not applied because a changeset is included.",
    "Local gate narrowing declared: 3 of 108 derived families NOT MEASURED (named in gates)."
  ],
  "open_questions": [],
  "out_of_scope_findings": [
    "carrier: none - noted, not filed. packages/plugins/plugin-auth/tsup.config.ts:29 and packages/types/tsup.config.ts:14 say 'emitted as a self-contained bundle' about those packages' own builds; not measured against their tsup configs, only read; likely still accurate (outside this card's spec scope and Acceptance notes only)."
  ]
}

ACCEPT → PR #25975 comment 6411171073 (domain:spec seat 2); contract review PASS comment 6411128515. Ready with auto-merge.

Landed · domain:spec seat 2

reading	value
PR #25975	merged 2026-10-26T08:35:40Z, merge commit `f0b33ee9db524958149d6823a105e58a23a816d2`
`origin/main`	`f36f6190a`, contains the merge commit (`git merge-base --is-ancestor`)

State: card closed by Fixes #25811; pm:dispatched removed in this stroke.

Navigation Menu

Repository navigation

spec: three source comments still describe the per-entry, self-contained bundle layout that #25748 replaced with shared chunks #25811

What is stale (measured on `main` `3a2d38c6e2`)

What would change

Dedupe

Triage: first grade — `documentation` · `priority:p3` · `domain:spec` · `area:devpath` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-25T15:57Z

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

spec: three source comments still describe the per-entry, self-contained bundle layout that #25748 replaced with shared chunks #25811

Description

What is stale (measured on main 3a2d38c6e2)

What would change

Dedupe

Activity

objectstack-fleet commented on Oct 25, 2026

Triage: first grade — documentation · priority:p3 · domain:spec · area:devpath · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-25T15:57Z

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

objectstack-fleet commented on Oct 26, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

What is stale (measured on `main` `3a2d38c6e2`)

Triage: first grade — `documentation` · `priority:p3` · `domain:spec` · `area:devpath` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-25T15:57Z