fix(spec): os migrate meta guidance for twenty-four more migration-entry families states each lesson in words, not tracker numbers (stage 8)

[objectstack-fleet]

Part of #20233
Stage 8: the stack-, evaluated-, aggregation-, authoring-, automation-, cache-, tenant-, client-, spec-, cli-, identity-, import-, tool-, advanced-, cloud-, startup-, sys-, declarative-, sort-, address-, packages-, platform-, session- and strategy- families.

Clause-②: no

os migrate meta prints each ADR-0087 semantic entry's surface, replacement, reason and acceptanceCriteria to the author. In these families, each sentence that cited a tracker number now says what the cited ruling, measurement or fix decided (form D). ADR ids stay, and so do the contributor-guide rule references, which are not tracker ids.

• Text only: the dev's base-vs-head AST comparison over the changed entry files and registry.ts found no change outside prose fields, except surface on the two entries whose surface carried an id.
• Pin: packages/cli/test/migrate-meta-engine-guidance.test.ts covers the 24 new prefixes; its ablation went red.
• Generated: registry.ts, spec-changes.json and docs/protocol-upgrade-guide.md, by their generators. patch changeset.

The measurements are in the dev's report on #20233 (5886017515).

Acceptance notes

• cli-command-contribution-retired said the manifest.contributes.commands tombstone was protocol 17. It is registered under 18, and the rewritten sentence says so.
• Comment and docblock lines of these entries still cite tracker ids. They are form C's, under packages/spec/src: 1,277 comment lines still cite 170 deleted tracker numbers (1,295 sites) — the staged remainder of ruling C+D on #19123, measured by PR #20226 #20234, and are untouched.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Sfe5YjBLwB9J3y8fvm2xq1

[github-actions]

📓 Docs Drift Check

2 anchor(s) derived from 1 changed package(s); no hand-written page names any of them. ⚠️ 1 changed file(s) yielded no anchor (packages/spec/spec-changes.json), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

What this run could not see

• 1 changed file(s) yielded no anchor (packages/spec/spec-changes.json) — pages documenting those are invisible to this run
• 4 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 54 of 206 client-bound route-ledger rows — the other 152 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 152: 0 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 55 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 97 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• 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 — 137 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 7001918e4955e2c1cad101b1f227199ebbf46966 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from 13dc659ff3426760d2da953ee8cbb350940c2034 — the merge of head 309e2f32714c992eade14ebd810996a58cd191bb into base 7001918e4955e2c1cad101b1f227199ebbf46966, 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 13dc659ff3426760d2da953ee8cbb350940c2034 && git checkout 13dc659ff3426760d2da953ee8cbb350940c2034
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 7001918e4955e2c1cad101b1f227199ebbf46966 309e2f32714c992eade14ebd810996a58cd191bb && git checkout -B drift-repro 7001918e4955e2c1cad101b1f227199ebbf46966 && git merge --no-ff 309e2f32714c992eade14ebd810996a58cd191bb

node scripts/docs-audit/affected-docs.mjs --json 7001918e4955e2c1cad101b1f227199ebbf46966

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

[objectstack-fleet]

Contract review

Served-tier: CONTRACT_REVIEW_TIER
Head-sha: 309e2f32714c992eade14ebd810996a58cd191bb
Local-runs: none

① Derived judgments

Inputs. Card #20233 body and all 30 comments (triage 5856629820; rulings carried — C+D on #19123 5749154545 form D, A on surface 5858839916; the stage 1–7 records; the stage-8 claim 5883700046 and os-dev-report 5886017515). PR #20607 body, its 34-file list, and the diff from the merge base f11b5f2 to the head (origin/main has since moved to eb4b17c; of this PR's paths only registry.ts moved there). The 85 distinct tracker ids the removed lines cited, read off the diff: 69 objectstack ids answering 200 and objectui#6812 by REST; the 12 answering 404 from main's commits and CHANGELOGs; cloud#2037, hotcrm#1141 and better-auth#10909 (403) from the objectstack records that quote them (#16526, #8687, #16629).

Accept-set and public-surface changes implied by the diff: none. The diff changes only the surface, replacement, reason and acceptanceCriteria strings of 29 semantic entries, their three regenerated projections, one test and one changeset. No schema, conversion, matcher, entry id, export or runtime string moves. Right: this is the prose-only change rulings C+D and A call for, and the shape stages 1–7 landed.

Prose only — checked on the merge-base..head pair, not taken from the report. (a) Every changed line in the 29 entry files sits inside one of the four prose string literals, read line by line. (b) Mechanically: with comment lines dropped, every string-literal body blanked and 'a' + 'b' runs collapsed, the token skeleton of each of the 30 changed .ts files under packages/spec/src/migrations/ is identical base vs head; the comment lines are identical; each file's id: values are identical; registry.ts carries the same 314 ids before and after. (c) Exactly two surface fields change — authoring-schemas-strict-unknown-keys and evaluated-expression-slots-source-required — both of which carried a tracker id at base, the only surface edits ruling A allows. spec-changes.json (20/20 lines) and docs/protocol-upgrade-guide.md (12/12 lines) change only that same prose; the generator gates ride Lint & Repo Gates, green on the head.

No tracker id left in author-shown fields of these families. On the head, the 32 entries the 24 prefixes select (29 changed plus the untouched client-envelope-convergence-analytics-automation, client-oauth-applications-delete-void, stack-config-default-export-unbuilt-refused) carry no # plus four or five digits and no repo#N form on any non-comment line. Three short #N tokens remain, all on lines this PR does not touch: AGENTS.md #0.1 (address-, replacement), Prime Directive #2 (evaluated-, reason), PD #10 (identity-, reason). None is a tracker id; PD #2 and PD #10 resolve to AGENTS.md's Prime Directives; #0.1 resolves to nothing in AGENTS.md today (the dev flagged it — ③). Comment and docblock lines still cite ids, unchanged, under #20234, as the PR body says.

Rewritten sentences vs their sources — what I checked and how I chose it. I read every rewritten line in the diff and resolved every one of the 85 distinct ids. Claim-level tests (the ruling comment or landing commit itself) went to every sentence whose replacement asserts a ruling date, an option / route / letter or a verbatim quote — #4721, #4936, #5073, #6409, #7630, #11825, #12776, #13802, #13881, #14478, #14788, #15136, #15811, #15939, #16059, #16325, #16421, #16526, #16629, #17667, #19543 (doors ① and ④), and the dated rulings carried on #6704, #8715, #10485, #14519. The paraphrased fixes and findings (#2922, #3464, #3715, #3950, #4001, #4095, #4212, #4286, #4666, #4834, #4988, #5055, #5143, #5499, #5581, #5907, #6037, #6048, #6083/#6279, #6203, #6350, #6486, #7300, #8075, #8421, #8687, #10194, #10724, #11811, #11833, #11925, #12036, #12209, #13388, #14238, #14902, #15430, #15479, #15807, #16186, #16293, #16296, #16320, objectui#6812) were tested against the card's title/body or main's commit. The "own card, id dropped" class (#4484, #5040, #5227, #5540, #5599, #5638, #6011, #6361, #6815, #12007, #13023, #17157) was tested only for the sentence staying true without the number, which holds wherever the entry id itself is named.

Findings by class:

• Surface rewrites — right. authoring-: "closed against undeclared keys by the unknown-key strictness wave" — 未知键静默剥离仍是全仓默认:把 #3405 的 strict 收紧从一个 schema 推广到整个可授权面(ADR-0078 完整性闸门) #4001 is that wave ([Decision] The #4001 unknown-key strictness wave is owed ADR-0087 entries — one per major, or one per batch? #7630's own title calls it so). evaluated-: "the 34 declaring positions of the census of engine-evaluated slots outside the flow ledger" — spec: the evaluated-slot rule of #15430 reaches only the flow-node ledger — every other ExpressionInputSchema slot an engine evaluates (formula expression, validation / hook / sharing condition, visibleWhen…) still accepts an ast-only or blank-source envelope #15811's title ("reaches only the flow-node ledger") and its census comment 5629834274 (36 positions; 34 after the two absorbed slots the reason explains); "the census figure of 36" right.
• The one correction — right. cli-command-contribution-retired: protocol 17 → 18. The tree carries retired-keys/18.kernel__Manifest__contributes.commands.ts, and be21955 registers the nine contributes members at major 18. This is the only sentence changed beyond id removal, as the PR body and changeset declare.
• Ruling dates, options, verbatims — all right. Two structured value classes (AddressSchema, LocationValueSchema) are all-optional stripping z.objects — a wrong key is accepted and dropped, so a stored-value scan reports a clean count it cannot earn #13802 A 2026-09-01 (5494663475). [finding] three sibling list doors declare limit/cursor and never read them, one reporting hasMore: false as a literal — REBUILD of #19365, which stopped resolving on 2026-09-21 #19543 door ④: 「退役，统一走 /meta/flow」 given 2026-09-25 "after asking why the flow list does not use the standard API" (5825819437, verbatim); door ① batch 🔗 Broken links detected in documentation #204 letter C 2026-09-21 (card body). [Decision] Does cloud re-create the four consumer-less control-plane schemas, or does step 3 just delete them? (follow-on to the #16325 ruling) #16526 A 2026-09-07 "cloud does not host the four consumer-less files; step 3 deletes them" (5567942216). [Decision] Should the cloud control-plane contracts leave @objectstack/spec and live in the cloud repo? (upstream of #12036) #16325 B "cut by owner" 2026-09-07, control-plane half moves, package/marketplace half stays. spec: duration-shaped number keys carry their unit in describe prose only — two ttl keys with different units in one block, bare timeout keys, unit-less tenant timeouts #14478 B 2026-09-02, "no grandfathered baseline" (5518649320). finding: check:duration-unit-keys reads .describe() but not JSDoc — a duration key documenting its unit only in JSDoc never enters the population, and one card already recorded a wrong reason because of it #15939 A 2026-09-11, the gate PR lands last after per-file fixes. [Decision] Per-user notification locale —— 2026-08-13 裁决所等的「实测拉力」已到(hotcrm:4 个已发布语言 × 16 个 notify 节点 × 0 可本地化) #13881 A 2026-09-01, sys_user.locale first-class once hotcrm#1185's measured pull arrived. spec: the kernel startup-orchestrator contract (IStartupOrchestrator, PluginStartupResultSchema) is declared, exported and documented, and implemented by nothing #16059 option 3 2026-09-06, re-declare as the shipped shape (5560223561). plugin-auth: decide the durable answer to better-auth's account-issuer rollback — 1.7.3 removed the issuer identity outright, and #16186 shipped an exact pin as a stopgap #16629 option 1, re-ruled 2026-09-10 on the four-facet analysis (5620179099) — the date the sentence keeps. AdvancedPluginLifecycleConfig is an authorable surface with no runtime consumer — nothing reads .health, and the kernel never constructs PluginHealthMonitor #11825 route 2 2026-08-25 (5404973907). [Decision] The #4001 unknown-key strictness wave is owed ADR-0087 entries — one per major, or one per batch? #7630 方向 A 2026-08-12. userFilters.allowAddTab 是 objectui 真读、spec 从未声明的活能力 —— promote 还是 reject?(挡住 #4001 批 18 的 UserFiltersSchema 收紧) #5073 A promote 2026-08-04. 声明式 apis:(ApiEndpoint)入站面全链路零执行:元数据装载成功、路由从未挂载、matchEndpoint 全仓无实现 #4936 v17 hard refusal of a non-empty apis: with build intent, 2026-08-04. 一个被静默丢弃的排序键对外部调用方仍然是静默的:direction 该 400 还是继续被丢掉(#4674 第 4 项) #4721 方案 A "两扇门同一个 change 关" 2026-08-03. drivers: implement count_distinct in the SQL family — enforcement half of the #6188 ruling #6409 the enforce half of the 2026-08-07 [spec] AggregationFunction 声明 8 个,SQL 族只实现 5 个 —— count_distinct / array_agg / string_agg 按 ADR-0049 enforce-or-remove 定去留 #6188 ruling (array_agg / string_agg retire, count_distinct stays and gets enforced). [finding] An accept-set narrowing owes a **BREAKING** banner in core but not in platform-objects — and the ADR-0087 classifier reads the banner #16421 B 2026-09-07 plus the attached "recorded — a row each in the ADR-0087 migration ledger" (5572145955). action.visible 的 current_user.positions 装的是 auth 角色而非安全层岗位,按岗位收敛的按钮对所有人静默消失(17.2.0) #15136 A 2026-09-05, batch Redesign homepage based on shadcn and fumadocs specifications #39 item 2. The /packages read doors' declared request schemas and their actual query reads diverge in BOTH directions — ?limit= and ?cursor= are declared and never read, ?type= is read and never declared #17667 route 2 2026-09-13, "a small bounded collection; pagination is not part of its meaning". [Decision] One spelling for the signed-in user's language on the published surface — SessionUserSchema.language (unproduced), sys_user.locale (#13881) and /auth/me/localization.locale (request/deployment) coexist #14788 D 2026-09-03. finding: StrategyContext.executeAggregate declares aggregations[].method as string while the engine contract declares the six-value AggregationFunction #12776: "one slot, one declaration" is the landing commit d028b37's own phrase.
• 404 ids — every rewrite sourced from main. 6037 → 1818998 (validate-only operation; "no hooks run"). 6048 → dca5bd3 ("ADR-0087 台账条目另行落地"). 6083/6279 → 53068c1 (ADR-0122 phase 2, bare names to the author state). 6704 → c3f4916 (ruling 2026-08-09, disposition A). 8715 → 2c86fe3 (2026-08-15, disposition B). 10194 → 2306a76 (theme validated at the /meta write door). 10485 → 35ad101 (ruled B 2026-08-21; /meta/theme gets the meta-unknown-type-namespace: a /meta type name that is not a plural of anything still mints a namespace — PUT /meta/fieldz/x answers 200 #8421 refusal). 10724 → be21955 (nine members, major 18). 12209 → service-analytics CHANGELOG ("a custom-SQL measure … already refused earlier, caller-facing, by resolveMeasureAggregation"). 14902 → 61821e5 ("Closes driver-sql: a PLAIN unique index (no organization key part) over existing duplicate rows fails the boot with the raw driver error naming no rows and no remedy, and os migrate plan classifies the same op safe with no duplicate pre-flight #14902"; a plain unique made loud and non-fatal), the MySQL hash-shadow arm being driver-sql: the MySQL hash-shadow arm still takes the boot down on a PLAIN unique over duplicate rows — the same guard #14902 moved on the direct arm #15479 (closed completed). 6361 dropped where the sentence names notification-list-cursor-retired, which the registry carries.
• 403 ids — right. cloud#2037: the id is dropped and "re-declared in the cloud repo beside their producer" stays, which [Decision] Does cloud re-create the four consumer-less control-plane schemas, or does step 3 just delete them? (follow-on to the #16325 ruling) #16526's body confirms. hotcrm#1141: "a downstream application's report of a top-level typo that shipped an artifact minus a whole family with validate and build both green" is 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's body (Part of hotcrm#1141; "drops an entire family out of the build with a green exit code"). better-auth#10909: the sentence already stated what 1.7.3 removed.
• Paraphrased fixes — all sourced. lifecycle.mdx's healthChecks TypeScript example does not match PluginHealthCheckSchema — the real config is checkMethod #11811 → lifecycle.mdx on main shows plugin[checkMethod] and monitor.registerPlugin(...) and states there is no healthChecks member. finding: two consumer-local structural DataEngineLike re-declarations remain after the #11493 deletion (datasource-admin-plugin, service-analytics) #11833 → c8be110 ("parsed with the spec's own enum, so a method the engine contract does not declare is refused"). callData 的 delete 成功体两条路径不同形状:protocol 回 {success:true}(合规范),ObjectQL 兜底回 {deleted:true}(spec 未声明的键) #5581 → 7bf3d1c (the fallback returns the declared success). bug(plugin-auth): published 17.1.0/17.2.0/17.3.0 float @better-auth/core to 1.7.3, which dropped createLocalAccountIssuer — a fresh objectstack dev --seed-admin never creates the system tables and never seeds #16186 → 2e6a2ea (exact 1.7.2 pin) after bug(plugin-auth): published 17.1.0/17.2.0/17.3.0 float @better-auth/core to 1.7.3, which dropped createLocalAccountIssuer — a fresh objectstack dev --seed-admin never creates the system tables and never seeds #16186's fresh-seed failure. [security] Two more spec schemas permit inline credentials: ExternalDataSourceSchema.authentication (clientSecret/apiKey) and MessageQueueConfigSchema.sasl.password — census toward #7990's parked boundary-guard reopen trigger #8075 → fork (b), "no sys_metadata door reaches either" (5270164082). finding(examples): the showcase seed writes f_address.postal_code, a key the value contract does not declare — accepted-and-stripped by the spec, and the ZIP box renders empty #13388 → "filed by the objectui#6812 survey seat while counting stored address values". [P2] Execute ADR-0033's deletion of the unenforced tool.requiresConfirmation (safety-shaped, gated by nothing) #3715 → "Execute ADR-0033's deletion of the unenforced tool.requiresConfirmation". spec: define the confirmation contract behind action.ai.requiresConfirmation — an unconfirmed run_action on a gated action is refused, and the refusal carries the way to confirm #16293 → the confirmation contract. The rest each state the cited card's title or body finding.
• Imprecise but not false. The two spec: duration-shaped number keys carry their unit in describe prose only — two ttl keys with different units in one block, bare timeout keys, unit-less tenant timeouts #14478 glosses — startup- ("the unit lives in the key name") and tenant-timeouts- ("carries its unit in its name") — state only the key-name arm of ruling B, which reads "in the key name or in the value". For these bare number keys the key-name arm is the operative one and the prescriptions (rename to *Seconds) are unchanged, so the author is not misled.
• No sentence found false, unsourced or over-broad. ADR ids stay throughout (ADR-0033/0049/0068/0078/0087/0090/0106/0122 references all present after the rewrite).

Pin. packages/cli/test/migrate-meta-engine-guidance.test.ts: COVERED_PREFIXES gains exactly the 24 prefixes the PR names; on the head registry each selects only its own family (32 entries; cli- does not reach client-, sys- not system-, packages- not package-), so the pin's own prefix rule holds; REWRITTEN gains the 29 changed ids (174 → 203). The assertion locates every selected entry's printed block verbatim in os migrate meta output and refuses # plus four or five digits in it, so the pin covers what the PR says it covers. The ablation is the dev's claim and was not re-run; Test Core and all six shards are green on the head.

PR body claims. "Text only … except surface on the two entries" — right. "Pin covers the 24 new prefixes" — right. "Generated … by their generators" — the generated diffs carry only the entry text and the generator gates are green. "ADR ids stay, and so do the contributor-guide rule references, which are not tracker ids" — right for PD #2 and PD #10; #0.1 is not a tracker id but points at nothing (③).

② Semver level

.changeset/20233-stage-8-migration-guidance-tracker-free.md: '@objectstack/spec': patch, a standalone Clause-②: no line, no widening / narrowing arm, no model identifier; the body names the two surface rewrites and the one correction. The diff publishes only from packages/spec (entries, registry.ts, spec-changes.json); the CLI change is a test file and docs/ is unpublished, so no second changeset is owed. The fix is to author-shown text of a released package (the card is a bug), which AGENTS.md's Post-Task Checklist step 3 assigns a patch changeset, never skip-changeset. No accept set moves, so Clause-②: no is right and no ADR-0087 disposition marker is owed. The PR body carries the same Clause-②: no. Check Changeset success. Level right.

③ Boundary flags

Dev report 5886017515 — open_questions: none.

deviations, each answered:

1. Deletion-leg ordering (run from base in its own worktree while the rewrite was already committed) — process only; its two reds were the tool pin's fragment assertions, and every asserted fragment (ai.requiresConfirmation, ACTION_CONFIRMATION_REQUIRED, 428, confirm: true) is in the head text. No effect on the diff.
2. Census and astdiff scripts rewritten in-container and validated against the stage 5–7 records — instrument provenance; my own skeleton compare in ① reaches the same text-only verdict independently.
3. The cli- protocol 17 → 18 correction — verified true on the tree (①) and declared in the changeset and PR body as required.
4. The pre-push hook refused a Part of line in the commit message; the commit was amended before any push — the relation lives in the PR body ("Part of os migrate meta prints tracker numbers to the author: ADR-0087 migration entries' reason / replacement / acceptanceCriteria text carries ~2,060 of them, 178 dead, which AGENTS.md's runtime-string rule forbids #20233"), the head is the branch's only commit, and Part-of PR must not also close its card and The card this PR closes must claim this branch are green.
5. Attribution — the head commit ends with the model-free trailer pair AGENTS.md prescribes (the session link and the generic co-author line), and the PR body ends with the session-URL footer; no model identifier in the changeset, the entries or the PR. Fine.
6. No labels or assignees written by the dev; bots set documentation / size/l / tests / tooling; needs:contract-review is on the PR and the card. Fine.
7. test/migrate-meta.e2e.test.ts NOT MEASURED locally (nightly by name) — declared; it is outside the merge queue's required set and no head check depends on it.
8. origin/main not merged; the dev's driver-free merge-tree is clean against eb4b17c with registry.ts the only moved path, and the merged registry equals regeneration — the seat regenerates on landing, as the claim's concurrency note already provides. GitHub reads the PR mergeable.
9. Teardown — no bearing on the diff.

out_of_scope_findings:

1. 76 prose-field sites remain in 39 entries, turso- (8) the largest — carrier: this card's later stages, as staged. Answered.
2. 832 comment / docblock lines still cite ids — carrier packages/spec/src: 1,277 comment lines still cite 170 deleted tracker numbers (1,295 sites) — the staged remainder of ruling C+D on #19123, measured by PR #20226 #20234 (form C), untouched here as the claim requires. Answered.
3. address-location-value-unknown-keys-refused's replacement cites AGENTS.md #0.1, and AGENTS.md carries no such numbering — 承接者:无. Escalated. It is not a tracker id, so it is outside this card's property, and the line is outside this PR's diff; it does not block this stage. It is nonetheless author-shown text pointing at a rule that does not exist by that number, and it should not stay noted only in a report: the seat should give it a carrier — a small card, or an explicit extra in the next stage of os migrate meta prints tracker numbers to the author: ADR-0087 migration entries' reason / replacement / acceptanceCriteria text carries ~2,060 of them, 178 dead, which AGENTS.md's runtime-string rule forbids #20233.

Claim 5883700046 stop-on-breach conditions, checked: no id / from / to / matcher change (①); no #20234 comment line touched (comment lines identical); no #20513 runtime string touched (nothing outside the entries, their projections, the pin and the changeset changed); none of the fenced open-PR entries (turso-, cube-, filter-is-empty-lowers-to-empty-operator, realtime-event-type-unemitted-values-retired) is in the file list.

Check-runs on the head 309e2f3, read last: 46 runs deduped to 35 names by newest started_at (newest 2026-09-29T07:56:38Z) — 30 success, 5 skipped (Auto Label, Build Docs, Check PR Size, Console Pin Gate, Packed-tarball smoke (opt-in)), 0 failed, 0 still running. Green: Build Core, Check Changeset, Lint & Repo Gates, Spec property liveness, Governed Surface Queue Guard, Test Core and its six shards, TypeScript Type Check and the four Type Check jobs, Dogfood Regression Gate and its three shards, Dogfood Verify CLI, Temporal Conformance (live PG + MySQL), the three card / branch guards, Check Documentation Links, Flag docs affected by code changes, filter. Commit status: one context (Vercel, ignored build step), success.

Implemented-by: claude/issue-20233-migrate-meta-tracker-free-stage-8
Reviewed-by: session_01Sfe5YjBLwB9J3y8fvm2xq1

VERDICT: PASS

Adopted and posted by domain:spec seat 5 (session_01Sfe5YjBLwB9J3y8fvm2xq1) · 2026-09-29T08:13Z · rendered by the seat's at-tier review subagent on this head; its served tier family was read from the subagent transcript before posting. ③'s escalated item, AGENTS.md #0.1 in the address- entry's replacement (not a tracker id; the seat found no such numbering in AGENTS.md), is carried by #20233 as an explicit extra of its next stage. PR #20610 landed on registry.ts after this head; a driver-free git merge-tree of this head onto origin/main 7510663c87 exits 0, and the merge queue runs CI on the merged state. Landing waits for all checks green.

---

Generated by Claude Code