[objectstack-fleet]

Requested by Steve · project thread
Fixes #31849

Clause-②: no
Metadata-protocol: none

维护者速读

• 改了什么：“这个对象是不是联邦对象（ADR-0015 external）”在运行时只剩一个函数：core 的 isFederatedObject，它直接读 spec 的 platformProvisionsStorage；plugin-security、plugin-sharing 改为转出 core 的同一个函数，另外 7 处手写的 external 判断也改为调用它。
• 为什么改：同一个判断写了十来遍，将来“联邦”的含义一变（比如改读 schemaMode）就得逐处去找，漏一处就会出现两层答案不一致。
• 风险与回滚：行为不变：对所有对象记录以及 null、undefined、数组、原始值，新旧答案相同。回滚即撤本 PR。
• 先看：packages/core/src/security/tenant-field.ts（isFederatedObject 本体）；packages/plugins/plugin-security/src/federated-phantom-anchors.ts（删掉的副本）；packages/core/src/security/tenant-field.test.ts（新钉）。
• 两分钟验证：git grep -n "external != null\|external == null" packages/core packages/objectql packages/plugins 预期在非测试源码里零命中。
• 需要你做的：无，席位落地。

What changes

• @objectstack/core isFederatedObject (security/tenant-field.ts) returns !platformProvisionsStorage(schema) from @objectstack/spec/data, the spelling [Decision] applySystemFields injects platform anchors into external objects the platform provisions no storage for — three consumers have now independently re-derived "that column is not really there" #7865's convergence map names.
• @objectstack/plugin-security and @objectstack/plugin-sharing federated-phantom-anchors.ts drop their own copy (security: external != null; sharing: its own !platformProvisionsStorage wrapper) and export { isFederatedObject } from @objectstack/core. The export name stays where it was, so every importer keeps resolving. @objectstack/objectql keeps re-exporting the core binding, unchanged.
• The other runtime copies of the presence test, found by the measure below, now ask the same binding:
  • packages/objectql/src/search-companion.ts isDdlManaged (external == null)
  • packages/cli/src/commands/migrate/unmapped-columns.ts (the "federated" outside reason)
  • packages/cli/src/utils/schema-migrate.ts (managed-drift driver scan)
  • packages/cli/src/utils/schema-migration-plugins.ts (coverage pass federated count)
  • packages/services/service-analytics/src/plugin.ts isExternalObject (ADR-0062 D6 gate)
  • packages/triggers/trigger-schedule/src/time-relative-trigger.ts organizationScopeIsInertFor
  • packages/metadata-protocol/src/protocol-save-door-refusals.ts (datasource-bound external read; was !== undefined && !== null)
• Comment-only rewording where a docblock named the old expression as the predicate: objectql engine-driver-options.ts, integrity/dangling-reference-audit.ts; plugin-security security-plugin.ts; plugin-sharing sharing-service.ts.
• One forced test edit: packages/objectql/src/federated-injected-column-readers.test.ts gains a READERS row search-companion.ts#isDdlManaged :: isFederatedObject() (not-a-read), because the search companion now calls the seam by name.

Why

Measure on origin/main 3f0e3ac (git grep -n "external != null\|external == null\|external !== undefined\|isFederatedObject\|platformProvisionsStorage" under packages/, non-test): besides the card's table, seven more source sites restated the presence test (listed above). Seat ruling 1 converges them too. Not converged, because they are a different question: service-datasource external-datasource-service.ts:1374 (external !== undefined || datasource !== 'default', a wider "routed off the default datasource" test) and datasource-admin-service.ts (patch.external, a patch field, not an object). packages/spec keeps the one definition (no spec edit).

Input equivalence (seat ruling 2): platformProvisionsStorage answers true (not federated) for null, undefined, arrays and any non-object, and the old ?.external != null answered false (not federated) on each of those, so the two agree. The only inputs where they differ are a function or an array that carries an own external property; no producer hands one to these seams (all read registry or overlay object records).

Tests

• packages/core/src/security/tenant-field.test.ts (new): isFederatedObject(x) === !platformProvisionsStorage(x) and the expected verdict for {}, { external: {} }, { external: null }, { external: undefined }, null, undefined, []; plus barrel identity.
• packages/plugins/plugin-security/src/federated-object-binding.test.ts and packages/plugins/plugin-sharing/src/federated-object-binding.test.ts (new): the module export toBe core's binding.
• Why identity: a restored local copy answers the same as core on every record, so no behavioural pin can see it (measured: the core ablation below stays green). Identity is the only thing that changes.

Ablations (ruling 4), each restored with blob == HEAD and empty diff

• plugin-security: delete the core import and restore export function isFederatedObject(schema) { return schema?.external != null } → federated-object-binding.test.ts RED (expected [Function isFederatedObject] to be [Function isFederatedObject] // Object.is equality), federated-tenant-layer0.test.ts stays green (behaviour identical). Tests 1 failed | 8 passed.
• plugin-sharing: delete the core import and restore the old !platformProvisionsStorage wrapper → federated-object-binding.test.ts RED, same assertion; Tests 1 failed | 57 passed.
• core: revert the body to ?.external != null → tenant-field.test.ts stays GREEN (8/8): the agreement pin cannot see a revert, by design, which is why the re-export layers carry the identity pin.
• Mutations landed through scripts/ablation-replace.mjs (anchor 1 → 0, marker grep 1); tests import the module under test from src, so no dist leg.

Acceptance notes

• Test files still name external != null in prose or as a test variant (e.g. injected-system-columns-parity.test.ts, the dangling-audit federated phantom test); they describe what a mutation would look like, not a predicate copy.
• packages/spec/src/data/injected-system-column-provenance.ts:263 still says plugin-security holds a hand-rolled copy; that doc is spec prose and this card forbids packages/spec/** edits.

Verification

Head 10eb82ed (origin/main 1a66fcf merged in; the merge brought #31876, which edits plugin-sharing and core in other files).

• git grep -n "external != null\|external == null" packages/core packages/objectql packages/plugins (non-test, CHANGELOG excluded): 0 hits. Under all of packages/, the only remaining hits are the definition in packages/spec/src/data/injected-system-column-provenance.ts.
• Build: turbo run build --filter='./packages/**' --filter='./examples/**' under the verify lock, VERDICT command-exit 0.

Package suites and typecheck

At 10eb82ed (after the main merge):

• core: typecheck + test, Test Files 141/141, Tests 3237/3237
• plugin-sharing: typecheck + test, Test Files 73/73, Tests 1497/1497
• plugin-security: typecheck + test, Test Files 252/252, Tests 6051 passed, 45 skipped
• objectql: typecheck + federated-injected-column-readers, federated-object, search-companion*, package-scoped* (6 files, 212/212)
• cli, metadata-protocol, service-analytics, trigger-schedule: typecheck, exit 0

At 73db2448 (before the merge; the code these suites exercise is unchanged since):

• objectql full: Test Files 580/580, Tests 11935/11935
• service-analytics full: Test Files 214/214, Tests 5140 passed, 332 skipped
• trigger-schedule full: Test Files 12/12, Tests 230/230
• metadata-protocol full: Test Files 330 passed, 5 skipped; Tests 22626 passed, 29 skipped
• cli --project unit: 404/406 on the first run. The 2 failures were published-subpath-console.pin and published-subpath-hook-body.pin, which need packages/cli/dist and stopped with "packages/cli is not built". After the full build they passed 2/2 (29 tests). The integration tier is left to CI.
• Lint: root eslint --no-inline-config on the changed .ts files, exit 0.

Derived gates

node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack at 10eb82ed: all 80 derived families exit 0. --ran reconciliation: "80 derived famil(ies) accounted for — 80 run, 0 NOT-MEASURED". node scripts/check-platform-checklist.mjs --changed-from origin/main: OK, 0 anchors lost.

Human Minimum

None; the seat lands it.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

This PR changes 8 package(s): @objectstack/cli, @objectstack/core, @objectstack/metadata-protocol, @objectstack/objectql, @objectstack/plugin-security, @objectstack/plugin-sharing, @objectstack/service-analytics, @objectstack/trigger-schedule, touching 11 documentable anchor(s). ⚠️ 1 changed file(s) yielded no anchor (packages/objectql/src/engine-driver-options.ts), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

3 hand-written doc(s) NAME something this change touched and may need an implementation-accuracy re-verification:

• content/docs/data-modeling/queries.mdx (via os migrate unmapped-columns (command, read off packages/cli/src/commands/migrate/unmapped-columns.ts))
• content/docs/deployment/cli.mdx (via os migrate unmapped-columns (command, read off packages/cli/src/commands/migrate/unmapped-columns.ts))
• content/docs/plugins/packages.mdx (via AnalyticsServicePlugin (symbol, a top-level class))

⛔ 2 release-owned page(s) also name something this change touched. These are read-only:

• content/docs/releases/v17/17-6.mdx (via AnalyticsServicePlugin (symbol, a top-level class))
• content/docs/releases/v17/17-7.mdx (via os migrate unmapped-columns (command, read off packages/cli/src/commands/migrate/unmapped-columns.ts))

content/docs/releases/ is RELEASE-OWNED (AGENTS.md "Documentation Guardrails"): release
notes are written centrally at release time, and a code PR that edits them is the exact PR
that guardrail exists to stop. They are still audited — read-only. If one of them is actually
wrong, file an issue or open a dedicated docs-only PR; do not edit it here.

What this run could not see

• 1 changed file(s) yielded no anchor (packages/objectql/src/engine-driver-options.ts) — pages documenting those are invisible to this run
• 1 name(s) were too generic to anchor anything (single lowercase words)
• the SDK route bridge reached 65 of 231 client-bound route-ledger rows — the other 166 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 166: 0 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 59 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 107 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 — 85 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 ac8fede8f938439d2228f5b1404042d113a3812e → packageMentionDocs.

Which tree this was computed on

This run read content/docs from 2e68567ae236e7502a78fc600a3931f3de960c02 — the merge of head 10eb82edc405e735543d18b08e21f1ec1dee51c7 into base ac8fede8f938439d2228f5b1404042d113a3812e, 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 2e68567ae236e7502a78fc600a3931f3de960c02 && git checkout 2e68567ae236e7502a78fc600a3931f3de960c02
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin ac8fede8f938439d2228f5b1404042d113a3812e 10eb82edc405e735543d18b08e21f1ec1dee51c7 && git checkout -B drift-repro ac8fede8f938439d2228f5b1404042d113a3812e && git merge --no-ff 10eb82edc405e735543d18b08e21f1ec1dee51c7

node scripts/docs-audit/affected-docs.mjs --json ac8fede8f938439d2228f5b1404042d113a3812e

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

Advisory only, and a precision-first one (#9192): a page is listed because it names a
symbol, wire route or SDK method this diff touched — not because it mentions a changed
package. Each row says which anchor put it there, so a wrong row is reportable rather than
merely annoying. To re-verify, run the docs-accuracy-audit workflow scoped to these files:
node scripts/docs-audit/affected-docs.mjs ac8fede8f938439d2228f5b1404042d113a3812e → pass the list as
args.docs, on the commit named under Which tree this was computed on.

[objectstack-fleet]

domain:server#2 · os-steve · session_01ADHkmQKZzp4DLtqiCkZZRM

维护者速读

• 改了什么： "这个对象是不是联邦对象（表在外部数据库）"的判断，以前在 core、plugin-security、plugin-sharing 以及另外 7 处各写一遍 external != null；现在都读 @objectstack/spec/data 的 platformProvisionsStorage 一个判定（core 的 isFederatedObject 是它的取反，其他包转出或调用同一个函数）。
• 为什么改： 联邦对象的定义以后只需改 spec 一处，不会各处说法分叉。
• 风险与回滚： 行为不变，每一处的判断结果与原来相同。回滚即还原本 PR。
• 先看： packages/core/src/security/tenant-field.ts。
• 需要你做的： 无，由席位落地。
• 席位意见： ACCEPT。

席位核对 — ACCEPT · domain:server#2 · Head-sha 10eb82edc405e735543d18b08e21f1ec1dee51c7 · 2026-11-09T11:13Z

1. Shape: draft to main, Fixes #31849, Clause-②: no, Metadata-protocol: none, patch changesets. 19 files, +156/−36.
2. Delta rulings, checked against the diff:
   • One predicate: core's isFederatedObject is !platformProvisionsStorage(schema); plugin-security and plugin-sharing re-export core's binding from federated-phantom-anchors.ts under the same names, pinned by toBe identity.
   • Measured census: 7 more source copies found and converged (objectql search-companion, cli unmapped-columns / schema-migrate / schema-migration-plugins, service-analytics, trigger-schedule, the metadata-protocol save-door datasource binding). Two left alone because they ask a different question (external-datasource-service's non-default-datasource arm, and the admin patch.external): accepted.
   • Same answers: on a non-object input the spec answers "provisioned", so this answers "not federated", as the presence test did. The function/array-with-external divergence has no producer and is not pinned (ruled).
3. Gates:
   • Dev: core 7-input pin plus barrel identity; both identity ablations red and restored; 80/80 derived gates, 0 NOT-MEASURED; typecheck green on 8 packages; the affected suites green.
   • Check-runs on this head: 54 success, 11 skipped, 0 failed.
   • Governed surfaces: 0 of 19 paths on the register.
4. Sequencing: PR feat(objectql, cli, metadata-protocol, spec, runtime)!: a namespace never equals an object name at every door #31887 (seat 1, K) edits protocol-save-door-refusals.ts in other hunks; whichever lands second merges main.
5. Landing: pr_ready + automerge_enable.