[objectstack-fleet]

Requested by Warren · project thread
Fixes #25098

Clause-②: no
Metadata-protocol: none

维护者速读

• 改了什么：系统设置（Setup）的「应用」组去掉未过滤的「软件包」入口；完整软件包列表只留在 Studio → 总览 → 软件包。
• 为什么改：维护者裁决「二、软件包列表挪到 Studio ；甲」：完整列表是开发面，Setup 只保留应用市场过滤后的「已安装应用」入口。
• 风险与回滚：只删一个导航项及其四语种标签；页面本身没动，书签照常可开。回滚即还原本 PR。
• 先看：setup-nav.contributions.ts（删掉的那一行）、新 pin setup-packages-door.test.ts、changeset。
• 两分钟验证：pnpm --filter @objectstack/platform-objects exec vitest run src/apps/setup-packages-door.test.ts，预期 3 个用例全绿。
• 需要你做的：无，席位落地。

What changes

• packages/platform-objects/src/apps/setup-nav.contributions.ts: Setup's group_apps contribution drops nav_packages (developer:packages, unfiltered). The group comment now says where the full list lives (Studio) and which Setup door remains: nav_marketplace_installed, the marketplace-filtered "Installed Apps" entry that @objectstack/cloud-connection contributes wherever the marketplace capability is present. Studio's nav_packages (studio.app.ts, Overview → Packages) is untouched.
• Translations: apps.setup.navigation.nav_packages is removed from en, zh-CN, ja-JP and es-ES, and its recorded digest from the three *.source-hashes.ts tables. apps.studio.navigation.nav_packages stays in all four catalogs and all three hash tables.
• setup-packaged-actions-nav.test.ts: restated, not loosened. Its retirement control used Setup's nav_packages as proof the walk reads the real array; it now uses nav_packaged_actions, which the same file already requires to exist. Its comment and one test title no longer say "beside Packages".
• New pin setup-packages-door.test.ts: Setup's shell plus SETUP_NAV_CONTRIBUTIONS carry no developer:packages entry without a source filter (and no nav_packages id), and Studio's Overview still carries exactly one unfiltered developer:packages entry. The predicate has its own positive and negative controls.
• Changeset: patch for @objectstack/platform-objects.

Why

Maintainer ruling 甲 on #25098 (amends #23892 ruling B, item 1): the full package list is a development surface, so its unfiltered entry lives in Studio only, and Setup keeps the marketplace-filtered Installed Apps door. The blocker objectstack-ai/cloud#3138 is cleared: cloud PR #3488 registers that door in hosted tenant environments, so a hosted Setup keeps a way to installed packages.

Tests

• pnpm --filter @objectstack/platform-objects test (under the verify lock, after building the dependency closure): Test Files 110 passed (110), Tests 1510 passed (1510).
• pnpm --filter @objectstack/platform-objects typecheck: exit 0.
• Ablation, one leg plus restore, at commit 510f2898 via scripts/ablation-replace.mjs: re-inserting Setup's nav_packages line into setup-nav.contributions.ts turns Setup ships no unfiltered developer:packages entry red (expected [ { id: 'nav_packages', …(4) } ] to deeply equal []; 1 failed, 2 passed). Restore blob equals the HEAD blob (e8992d3f9d5b) and git diff HEAD is empty. The test imports the source file directly, so no dist/ rebuild sits in the path.

Acceptance notes

• packages/cli/scripts/check-app-nav-i18n.mjs names nav_packages only in two comments that record Setup 应用 4 条运行时贡献的导航在 zh 下仍是英文，而被两处注释同时指定的 coverage ratchet 报绿 —— 能力插件贡献的 nav 不在任何一次走查里 #5750's history (the id was untranslated at the time; the Studio copy did not answer for Setup). Both stay true as history, so they are unchanged. The gate judges the merged Setup app, which no longer carries the id, so it has one id fewer to check.
• packages/cloud-connection/src/marketplace-installed-ui-entry.pin.test.ts describes developer:packages as "the component Setup's own Packages entry opens" in a doc comment. That clause is now stale; it is outside this card's file surface and left for the seat.
• Setup's group_apps translations still list nav_cloud_connection; its move to Integrations belongs to cloud-connection: a runtime holding the env service key is platform-managed: register no Setup page, refuse unbind; move the entry to Integrations (ruling 甲) #25097, not this card.

Verification

All readings at head 510f2898.

Derived gates: 59 derived, 57 exit 0, 2 NOT MEASURED; plus pnpm lint and check:app-nav-i18n

• node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack derived 59 commands from the 11 changed paths (merge base 399df5bb).
• 57 exit 0. pnpm check:i18n and pnpm check:lean-entry-closure first exited 3 (prerequisite not built); after turbo run build of the closure the i18n gate names, plus @objectstack/objectql, both exit 0: check-i18n-bundles: OK (10 package(s) — all bundles in sync, no undeclared authoring keys).
• NOT MEASURED, reason: prerequisite not met (exit 3). pnpm check:dual-build-cjs-loads needs a full pnpm build of about 83 packages; CI's Build Core runs it. node scripts/check-plugin-teardown-shape.mjs --self-test is checker-health only and needs an unshallow clone.
• pnpm check:app-nav-i18n (outside the derivation, run because the card names it): OK (13 contributor(s), 4 locale(s), 2 app(s) — setup: 56 merged nav id(s), account: 12 merged nav id(s) — every id labelled in every locale).
• pnpm lint under the verify lock: VERDICT command-exit 0.
• dispatch-gates --ran: 59 derived famil(ies) accounted for — 57 run, 2 NOT-MEASURED (2 DERIVED from a recorded exit 3).
• CI families (type-check lanes, Build Core, Test Core) are read from this PR's check-runs, not re-run locally.

Human Minimum

• 1. Schemas start as Zod; packages/spec holds no runtime logic. (not touched)
• 2. Every name I added uses the casing for its kind.
• 3. No formatter ran; pnpm lint passes.
• 4. One issue, one branch off an up-to-date main, one pull request.
• 5. A change that publishes carries a changeset, and both declaration lines are filled in here and in the changeset.
• 6. packages/spec touched: its generated artifacts are regenerated and committed. (not touched)
• 7. No hand edits to content/docs/references/, packages/*/CHANGELOG.md or content/docs/releases/.
• 8. Title, body and comments are in English. (The maintainer quick-read section is in Chinese by the os-dev order.)

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

5 anchor(s) derived from 1 changed package(s); no hand-written page names any of them, so this run has nothing to list — not a clean bill of health. This check sees only pages that NAME a derived anchor: one that documents this change in prose, or enumerates it in an authoring dialect, names none and stays invisible to it on every run.

What this run could not see

• 1 name(s) were too generic to anchor anything (single lowercase words)
• 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 — 2 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 208d3e2efb47d21dc05793cae500d1d76ea25a21 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from a4079afda793478a377c217ae11b4f7d4b292795 — the merge of head b87abb91b4dfb8cf5e45ff6fe8838745bf39a325 into base 208d3e2efb47d21dc05793cae500d1d76ea25a21, 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 a4079afda793478a377c217ae11b4f7d4b292795 && git checkout a4079afda793478a377c217ae11b4f7d4b292795
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 208d3e2efb47d21dc05793cae500d1d76ea25a21 b87abb91b4dfb8cf5e45ff6fe8838745bf39a325 && git checkout -B drift-repro 208d3e2efb47d21dc05793cae500d1d76ea25a21 && git merge --no-ff b87abb91b4dfb8cf5e45ff6fe8838745bf39a325

node scripts/docs-audit/affected-docs.mjs --json 208d3e2efb47d21dc05793cae500d1d76ea25a21

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

[objectstack-fleet]

domain:server#4 · os-warren · session_01JQwxbacSXJQqAztgt4UMrS

席位核对

ACCEPT · domain:server#4 · head b87abb91b4dfb8cf5e45ff6fe8838745bf39a325 · 2026-11-02T19:37Z · seat review of record at TIER_DEFAULT (not governed: check-governed-merges --pr 29547 0 of 12 paths; Clause-②: no; no packages/spec path)

1. Shape: draft, base main, body line 3 Fixes #25098. 12 files, +116/-36: one patch changeset for @objectstack/platform-objects, Setup's group_apps loses nav_packages, the key and its digests leave the four catalogs and three hash tables (Studio's nav_packages stays in all of them), one restated test, one new pin, and the CLI composition count. Seat read the diff: Studio's Overview entry (studio.app.ts) is untouched, and Setup keeps the marketplace-filtered nav_marketplace_installed door.
2. Premise measured: ruling 甲 on Setup → Packages moves to Studio: drop Setup's nav_packages; Studio keeps the full list, Setup keeps the marketplace "Installed Apps" door (ruling 甲, amends #23892 B) #25098 holds on origin/main: Setup carried an unfiltered developer:packages entry beside Studio's. The blocker cloud#3138 is cleared (cloud PR feat(spec)!: prune dead ReportColumnSchema/ReportGroupingSchema exports + unread report chart groupBy (#3463) #3488, per the card). The new pin's ablation reds only the Setup case (1 failed / 2 passed), and the restore equals HEAD.
3. Gates on this head: check-runs pinned to b87abb91: 48 total, 35 success, 13 skipped, 0 failed; Lint & Repo Gates, TypeScript Type Check, Test Core (8/8 shards) and Build Core success. mergeable_state clean. Dev local at 510f2898: dispatch-gates --ran 59 derived / 57 run / 2 NOT MEASURED (check:dual-build-cjs-loads, covered by Build Core; teardown self-test, checker health only), pnpm lint 0, platform-objects 1510/1510.
4. Accepted deviations: head b87abb91 adds one file outside the card's surface, packages/cli/test/connect-agent-both-halves-wire.pin.test.ts: the Setup tree count goes 34 → 33, the direct result of the removed entry, with the reason written in its comment. It was red on Test Core (4/8) at 510f2898 and is green here. The body's Verification readings name 510f2898; this one test file is the only difference, and CI is read on the new head.
5. Out of scope: the stale doc-comment clause in packages/cloud-connection/src/marketplace-installed-ui-entry.pin.test.ts ("Setup's own Packages entry") is a test comment, not published text; the seat drops it rather than filing a card. nav_cloud_connection's move stays with cloud-connection: a runtime holding the env service key is platform-managed: register no Setup page, refuse unbind; move the entry to Integrations (ruling 甲) #25097. Skills-docs: n/a (Metadata-protocol: none).
6. Landing: not governed, under 3000 lines ⇒ the seat lands it (pr_ready + automerge_enable). At MERGED: landing record on Setup → Packages moves to Studio: drop Setup's nav_packages; Studio keeps the full list, Setup keeps the marketplace "Installed Apps" door (ruling 甲, amends #23892 B) #25098 (closed by Fixes), pm:dispatched removed.