[objectstack-fleet]

Part of #27114

Clause-②: no
Metadata-protocol: none

维护者速读

• 改了什么: 只改了 packages/cli/test/migrate-meta.e2e.test.ts 这一个夜间 e2e 测试文件。fixture 换成第 17 步仍注册的 12 个 conversion 各一处写法；os migrate meta --from <N> cannot load the retired-key sources it exists to fix — validates through the post-retirement schema before converting #9418 那组用例改用仍会被拒并被转换的退役写法：structuredOutput.format: 'regex'（agent-structured-output-refused-members-removed）和 dashboard.refreshInterval；不用 agent.lifecycle，因为它的退役测试钉死该键在全树不得出现。
• 为什么改: feat(spec)!: retirement kits whose conversion retired at or below 16.x age out; the strict unknown-key refusal names the protocol upgrade guide #26599 让 16.x 的 45 个 conversion 随退役套件老化离开第 17 步，--from 16 不再改写这些写法，快照里留下未知键，夜间 tiers 因此 6 个用例变红；PR 和合并队列不跑 e2e 层，所以只在夜间暴露。
• 风险与回滚: 只动测试，不动产品代码，不发布任何东西。用例数不变（19 个），没有跳过、改名或换层。回滚就是 revert 本 PR。
• 先看: ① 该文件里的 PRE17_CONFIG 和 EXPECTED_CONVERSIONS；② 新增的注册表钉子（EXPECTED_CONVERSIONS 必须等于 MIGRATIONS_BY_MAJOR[17].conversionIds）；③ os migrate meta --from <N> cannot load the retired-key sources it exists to fix — validates through the post-retirement schema before converting #9418 那组的 fixture 和 os validate 断言。
• 两分钟验证: 先建好 cli 的依赖闭包，再跑 OS_TEST_TIERS=nightly pnpm --filter @objectstack/cli exec vitest run test/migrate-meta.e2e.test.ts，预期 19 个全部通过。
• 需要你做的: 无，席位落地。

What changes

#26599 (dbf7e04adf, ruling 6434038284 on #25767) removed the 45 conversions retired at 16.1.0 from step 17's conversionIds. It re-pointed the unit-tier cli pins but not this file, which runs only under OS_TEST_TIERS=nightly. The 2026-10-28 nightly (run 40638682260) went red on six of its cases.

All six failures come from that narrowing:

• The main fixture carried one shape per 16.x conversion family. Nine of the ten families it expected are no longer registered, and --from 16 leaves those shapes as written. So the run attributes nothing to them, and the snapshot keeps responsive, performance, aria and defaultSort as unknown keys. That breaks schemaValid, the per-family attribution, the --out re-parse and idempotence.
• The os migrate meta --from <N> cannot load the retired-key sources it exists to fix — validates through the post-retirement schema before converting #9418 fixture carried agent.knowledge and skill.triggerPhrases. Their tombstones and conversions left the spec with the 16.x kits, so the load no longer converts them, and os validate answers with the strict unknown-key refusal that names the protocol upgrade guide, not the retired-key prescription.

The fix is in this one file. No case is added, removed, renamed or skipped, so the count stays at 19.

• Main fixture: the aged-out shapes are dropped (option (a) on the claim). The fixture now carries one shape for each of the 12 conversions step 17 still registers: the sharing-rule full → edit rename, the seven flow-node aliases and lifts (crud object, notify, wait, connector, map, subflow, script), mongo → mongodb, the converged retry policy (retryDelayMs → backoffMs), the page-header description → subtitle alias and the inline api action's object-form params → bodyExtra. The list was read from registry.ts.template step 17, not from memory, and every id was checked against conversions/registry.ts. The required / no-storage assertions (ADR-0087's field-required-notnull-explicit conversion asserts an implication ADR-0113 abolished — the boot calls it a forward conversion, but its output and the source it prescribes disagree at the storage layer #16693), the canonical delete_record pass-through and the flow errorHandling.maxRetries 有两个默认值:schema 声明 0、引擎回退 3 —— 重试次数取决于流程有没有过 schema #4247 maxRetries TODO stay.
• Registry pin: the attribution case first asserts that EXPECTED_CONVERSIONS equals MIGRATIONS_BY_MAJOR[17].conversionIds (imported from @objectstack/spec/migrations, as migrate-meta-default-range-window.test.ts does). The next family that leaves or joins the step then reds here by name, not as a missing rewrite.
• os migrate meta --from <N> cannot load the retired-key sources it exists to fix — validates through the post-retirement schema before converting #9418 fixture: re-pointed to retired spellings whose refusal and conversion still stand. defineAgent carries structuredOutput: { format: 'regex' }, a retired member refused at parse (agent-structured-output-refused-members-removed), and Dashboard.create from @objectstack/spec/ui carries refreshInterval (dashboard-refresh-interval-to-refresh-interval-seconds). Both are protocol 18, which --from 16 replays into. agent.lifecycle is not used, because its retirement pins the key absent from the tree. There are still two per-artifact helpers from spec subpaths, so a tolerance scoped to defineStack alone still fails it. The os validate half now matches the refusal's own prescription, --from 17.

Acceptance notes

1. Option (a), not (b). If an aged-out shape were kept, the replayed snapshot would be schema-invalid by design. That would break the schemaValid, re-parse and idempotence cases that the 12 surviving families need, and asserting the refusal instead would need a separate fixture and a new case. The ruled refusal itself, the unknown-key message naming the upgrade guide, was observed while reproducing this failure (the nightly log's os validate output). The spec owns the unit pins for it.
2. Wider family coverage than before. Before the narrowing, the file covered one step-17 family that still survives: sharing-rule. It now covers all 12 that step 17 registers, which restores the docblock's stated purpose: "one key from every v17 conversion family in a single pass".

Tests

At head e3006f2bd (the patch that moved the #9418 fixture off agent.lifecycle; addendum 6464184183 on #27114): packages/spec test:repo 81/81 files, 1324 tests (agent-lifecycle-retirement.test.ts 14/14); the nightly e2e 19/19; the cli unit project 343/343 files; derived gates 47 at exit 0, 1 NOT-MEASURED (check:dual-build-cjs-loads, dist absent). The readings below are the first head's.

At head f73c11c8 (this branch, merged with origin/main 6817bf1ea):

• OS_TEST_TIERS=nightly pnpm --filter @objectstack/cli exec vitest run test/migrate-meta.e2e.test.ts: before the fix, at d90657b2, Tests 6 failed | 13 passed (19), the same six the nightly names. After: Tests 19 passed (19), lock VERDICT command-exit 0.
• pnpm --filter @objectstack/cli exec vitest run --project unit: 341 of 343 files and 4929 tests passed, 29 skipped. The two remaining files (published-subpath-console.pin, published-subpath-hook-body.pin) refused before any test because cli's own dist/ was not built. After pnpm --filter @objectstack/cli build both passed (29 tests).

Verification

Derived gates (dispatch-gates --commands, 48 families)

• 47 exited 0.
• node scripts/check-plugin-teardown-shape.mjs --self-test first exited 3 (prerequisite: its pinned positive-control commit was outside the shallow clone). After fetching that commit, it exited 0 (48 cases).
• pnpm check:dual-build-cjs-loads exited 3, PREREQUISITE NOT MET: it reads every package's dist/, and only the cli closure was built. Recorded as NOT MEASURED; CI's Lint and Repo Gates runs it.
• dispatch-gates --ran: 48 derived, 47 run, 1 NOT-MEASURED (derived from the recorded exit 3), 0 UNRUN.

---

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 e6e41998a8d4543f5935e7c45145dfc3d6fe0e26 → packageMentionDocs.

[objectstack-fleet]

席位核对 — ACCEPT · domain:devx#2 · head e3006f2bdc · 2026-10-28T08:26Z

席位意见：合入。修掉 #27114 的第一个失败文件（cli 的 migrate-meta e2e）：fixture 对齐第 17 步仍登记的 12 个 conversion，并加了注册表钉子，以后再有 conversion 离开第 17 步会直接点名报红；测试一个没删没跳过。#27114 还有第二个失败文件（showcase live），所以正文是 Part of，卡继续开着。

Shape: → main, Part of #27114. The seat edited this through the relay (run 40656668108, read back identical) from Fixes, after the second failing file surfaced (reading 6463483048). The body has no other closing keyword. Label skip-changeset, assignee os-justin. 1 file, packages/cli/test/migrate-meta.e2e.test.ts, +111/−106. Reports: os-dev-report 6463368173, addendum 6464184183, correction 6464195356.

Seat's own read:

• The cause matches feat(spec)!: retirement kits whose conversion retired at or below 16.x age out; the strict unknown-key refusal names the protocol upgrade guide #26599 (dbf7e04adf). Step 17 lost the 45 aged-out ids, and the old fixture still carried their shapes.
• The new fixture carries one shape for each of the 12 ids step 17 registers. Each has a spot-check on the snapshot.
• EXPECTED_CONVERSIONS is pinned equal to MIGRATIONS_BY_MAJOR[17].conversionIds.
• The os migrate meta --from <N> cannot load the retired-key sources it exists to fix — validates through the post-retirement schema before converting #9418 pair moved off agent.lifecycle, whose absence pin agent-lifecycle-retirement.test.ts reddened the first head. It now uses structuredOutput.format: 'regex' (agent-structured-output-refused-members-removed, registry L9677 on main) and dashboard.refreshInterval.
• The absence test itself is untouched.

First head red, owned and fixed: Test Core (2/8) on f73c11c812, packages/spec test:repo. The cause was this PR's own fixture. The patch round turned it green on e3006f2bdc.

Gates: at head e3006f2bdc, 51 check-runs: 34 success, 17 skipped, 0 failure. Test Core 1–8/8 and PR-body guards re-ran after the body edit; both are success. check-expected-skips --pr 27132: OK (17 skips, all in main's roster). check-governed-merges: not governed, 217 lines.

Owed after landing (on #27114): the next scheduled nightly is the first that can show both files. The cli file should be green. The showcase file is claimed separately: claim 6463869260, cloud session session_01QSeKYobVx7yTf1SUMMbKnP.

Landing: non-governed. Ready + auto-merge now.

---

Generated by Claude Code