chore(objectui): bump the console pin to 2e818d0b51ec (carries objectui#11466, #11574, #11578, #11581 and #11583)

[objectstack-fleet]

Fixes #21696
Clause-②: no

This moves the bundled Console's objectui pin from ab1879721595 (#21625) to objectui main as of the write time, 2e818d0b51ecdf8fdd9fcbf4b916bcd7fe9a9cf6, which is past 973fc20f3. The Console now carries objectui#11574 (the phantom "(empty)" groups fix), objectui#11578, objectui#11581, objectui#11576, objectui#11583 and objectui#11466 (V1 declared node slot). It also carries objectui's own 17.7.0 package release.

⛔ This is not a release act. Version PR #21352 is untouched, and no tag, publish or Release was made. Whether 17.7.0 ships with this pin is the maintainer's call.

Range

ab1879721595..2e818d0b51ec has 22 non-merge commits. git merge-base --is-ancestor exits 0 against objectui origin/main for 973fc20f3, and 2e818d0b51ec was objectui main when it was read (git ls-remote just before this PR).

From the changesets objectui declared over the range:

• 17 releasing changesets, 1 release-nothing, and 6 commits that carry no changeset.
• 12 of the 18 changesets were already consumed by objectui's release b493919c7. Each was read from the commit that added it.
• 2 breaking entries, both annotated by the author and both from one commit.

The range was listed by measurement, not copied from the card. Exactly one commit subject in the range carries !:

git log --format='%h %s' ab187972..2e818d0b5 | grep -E '^[0-9a-f]+ [a-z]+(\([^)]*\))?!:' gives 83e3f8377 feat(types,react,core,plugin-dashboard)!: … (objectui#11466).

objectui landing	commit	note
objectui#11466: node slot and SchemaRenderer's schema take DeclaredNode; the dashboard producers name declared nodes; an envelope's object-metric retires	83e3f8377	declared breaking, both breaking entries
objectui#11574: a grouping-field switch paints only the server's groups	2f54dca53	smoke below
objectui#11581 / #11578 / #11576: Create View doors share one builder; a refused save is said; chart views bind an ADR-0021 dataset	b65aa5e65, f1c937966, d0097af2e
objectui#11583: a refused metadata write on the view, report and draft surfaces is said	e8c0b9614	past the card's 973fc20f3
objectui#11598: dashboard slot-entry and chart-producer reads compile without BaseSchema's index signature	2e818d0b5 (the pin itself)
objectui#11572 / #10380: interface and object page relay stored view config	09036173c, 068564691
objectui#11588: grid summary footer reads currency/precision from the object field only	d768c3178
objectui#11216: object-kanban declares grouping as the spec's GroupingConfig	a7557a7d4
objectui#11573, #11564: AnyComponentSchema / flex child typing	bdc9049ed, b10c68e58
objectui release 17.7.0	b493919c7	removes 2726 consumed .changeset/*.md files

ADR-0087 disposition of the declared-breaking entries

Both entries are not-required (no-migration-prescription). The disposition is in .changeset/console-2e818d0b51ec.md, which replaces the script's adr-0087: TODO placeholder.

objectui PR	commit	entry	disposition
objectui#11466 (PR objectui#11512)	83e3f8377	A node slot and SchemaRenderer's schema prop take DeclaredNode	not-required (no-migration-prescription). This is objectui's own TypeScript face (@object-ui/types / @object-ui/react). No ObjectStack schema declares or references it.
objectui#11466 (PR objectui#11512)	83e3f8377	An object-metric node in a dashboard widget's legacy component envelope draws the retired-format prompt	not-required (no-migration-prescription). An ObjectStack author cannot write that envelope: ui/dashboard.zod.ts refuses a widget's component key by name as objectui-internal, and no example authors one. The showcase's object-metric page tiles still draw their numbers (smoke below).

check-adr-0087-registration --base origin/main gives "1 declared-breaking changeset(s), each carrying an ADR-0087 disposition". check-changeset-no-major --base origin/main gives "This diff introduces no major bump".

What changed here

• .objectui-sha and .changeset/console-2e818d0b51ec.md were written by scripts/bump-objectui.sh 2e818d0b51ecdf8fdd9fcbf4b916bcd7fe9a9cf6 --no-commit with OBJECTUI_ROOT set to a read-only clone of objectui. The level was auto-set to minor. No ../objectui sibling exists in this container.
• scripts/sdui-manifest.record.json was re-recorded by node scripts/gen-sdui-manifest-node.mjs over the tree that pnpm objectui:build built at the pin. sdui.manifest.json is byte-identical: 107 components, sha256 0ead67c1111d… at both pins, so no component input moved. The record moves its pin, its modulesRoot and objectui's workspace version (17.6.0 to 17.7.0).
• packages/sdui-parser/objectui-lockstep.json was re-recorded with pnpm gen:sdui-lockstep against OBJECTUI_ROOT=.cache/objectui-2e818d0b51ec. It records 214 grammar lines (blob 0131f27cf86d), 25 codes and containment predicate 76c18fb95d1f. All are unchanged, so no port is owed. objectui's packages/sdui-parser/src is byte-identical across the range; only its CHANGELOG and package.json moved.
• The 54 asserting pin citations in packages/spec/src were re-measured at the new pin, not restamped:
  • Every anchor in each asserting record was resolved against objectui's tree at the old pin and mapped through git diff -U0 ab187972 2e818d0b5.
  • Four cited files changed:
    • plugin-dashboard/src/index.tsx (+13/-1, objectui#11466). Lines were added above the object-metric registration, so its start moves 244 to 256 and the icon input moves 269 to 281. Both lines are byte-identical, still { name: 'icon', type: 'string' }.
    • packages/types/src/objectql.ts. objectui#11216's grouping member landed below the cited limit member, so :4720 did not move.
    • content/docs/plugins/plugin-kanban.mdx. A grouping row landed below the limit row, and the page still teaches limit: 250.
    • packages/react/src/SchemaRenderer.tsx (+21/-20, objectui#11466). Only the type import, the SchemaRendererProps docblock and the schema type changed. The properties.* hoist, its strip list and the createElement spread did not, so object-timeline's prop channel still carries data.
  • Every other cited file is byte-identical across the hop.
  • Each record gains a dated 2026-10-04 hop sentence and keeps its ab1879721 history.
• The six migration entries' corpus counts were re-taken with git grep -o -F. That method first reproduced every ab1879721 number: 10267 files, objectstack 16377, @objectstack/spec 6665, timeout 1348, useState 2476, TTL 180, tenant 1238, Span 491.
  • The new numbers are 7579 files, 17227, 7134, 1351, 2477, 182, 1317 and 505. The file count drops because objectui's 17.7.0 release removed 2726 consumed changesets.
  • Every zero is still zero: 98 tokens were checked, the export lists of the three cited spec files plus every named key. The only non-zero tokens are the expected Span / SpanSchema. The 14 new Span hits are Spanish prose.
  • packages/spec/src/migrations/registry.ts was regenerated with gen:migration-registry.
• .changeset/objectui-pin-citations-2e818d0b51ec.md is a @objectstack/spec patch, because the FormField.span describe and six migration descriptions name the pin. content/docs/references/ui/view.mdx was regenerated by check:generated --fix.
• main was merged (7e0066af7a) with scripts/pm/os-regen-merge.sh in merge commit a0a35cb30f, with no conflict hunk. Afterwards:
  • check:generated reports all 15 artifacts current.
  • main's three new semantic entries (ui-object-form-fields-names-typed, ui-object-gantt-markers-typed, ui-object-timeline-mapping-typed) are present in the regenerated registry.
  • The pin-citation gate still counts 54 asserting citations. main added historical ones only (108 to 118).

No example, test or gate needed adapting, and no code changed outside generated records, citations and changesets.

Browser smoke: examples/app-showcase with the Console built at 2e818d0b51ec

The Console was built by pnpm objectui:build. The build log reports: "Console dist matches the objectui pin (objectui@2e818d0b51ec)", "Single-zod canary: exactly one zod version literal {major:4,minor:6,patch:5}" and "Console bundle carries THIS tree's @objectstack/spec, and only it".

The server ran pnpm dev -- --fresh --ui --no-watch -p 41853 (with OS_PORT=41853) on its own ephemeral DB with the seeded admin. Headless Chromium (/opt/pw-browsers/chromium) drove it, with console messages, page errors and every non-2xx response captured. Only the PIDs this run started were stopped.

Dashboards. objectui#11466 touches the dashboard producers. There are no page errors and no "retired" prompt on any of the four dashboards.

dashboard	metric tiles (value read in the DOM)	charts drawn	console errors
Delivery Operations showcase_ops_dashboard	Active Projects 2, At-Risk (Red) 1, Awaiting Review 2, Total Budget 1,090,000	4	404 /api/v1/usage/storage, 404 /favicon.ico
Revenue Pulse showcase_revenue_pulse	Invoices 7, Invoiced Subtotal 1,920, Accounts Signed 3, Accounts (all time) 14, Paid Rate 42.9%	4	404 /api/v1/usage/storage
Chart Gallery showcase_chart_gallery	Total Tasks 10, Avg Progress 47.0, Total Spent 616,000	13	404 /api/v1/usage/storage
System Overview system_overview (setup)	Total Users 3, Active Sessions 3, Login Events 3, Config Changes 0	2	404 /api/v1/usage/storage

The Delivery Operations tiles match the REST data on the same server:

• showcase_project?status=active returns 2 rows, ?health=red returns 1, and the budget sum is 1,090,000.
• showcase_task?status=in_review returns 2 rows.
• showcase_task has 10 rows, matching the Chart Gallery's Total Tasks.
• showcase_account has 14 rows, matching Revenue Pulse's all-time count.

The object-metric page tiles that this repo authors also draw:

• My Work shows Open Tasks 8, In Review 2 and At-Risk Projects 1.
• Command Center (大屏) shows 2 / 8 / 2 / 1 / 14 / 1.1M.

Neither page has a page error.

Grouping-field switch (objectui#11574). The drive was Tasks list, then Group, then Add group field (it selects Title). The select then switched to Priority, Status and Priority. The painted groups are counted from the DOM, and the server answers are each POST /api/v1/data/showcase_task/query with groupBy, all 200:

step	painted groups	server rows	"(empty)" groups	stuck "Loading grid…"
Title	10 (one per task)	10	0	0
Priority	4: High 3, Low 1, Medium 4, Urgent 2	4	0	0
Status	5: Backlog 2, Done 2, In Progress 2, In Review 2, To Do 2	5	0	0
Priority	4: High 3, Low 1, Medium 4, Urgent 2	4	0	0

The same sequence painted 13, 8 and 8 groups (9, 3 and 4 phantom) at ab1879721595, as #21625's acceptance notes recorded. No phantom "(empty)" groups appear at this pin.

Console errors across the run. The only ones are the browser's "Failed to load resource: 404" lines for /api/v1/usage/storage and /favicon.ico, and the run has 0 page errors. The usage/storage 404 was already recorded in #21625's smoke.

#21671 lockstep reading

• objectui's packages/sdui-parser/src/validate.ts at 2e818d0b5 is byte-identical to ab1879721595.
• checkType (:450-466) still grades type-mismatch with severity: arms.includes('enum') ? 'error' : 'warning'. checkMemberTypes (:407-419) grades member-type-mismatch the same way.
• This repo grades both error since PR fix(sdui-parser): an html page literal of the wrong type is a compile error, not a warning #21678 (packages/sdui-parser/src/validate.ts:446, :506). The gap still stands at this pin.
• This repo's "Known lead" note (packages/sdui-parser/src/validate.ts:336-340) names objectui's copy "at the .objectui-sha pin" without a sha, so it is still true and needs no update. The bump route regenerates nothing there, and validate.ts is not edited.
• check:sdui-lockstep does not compare severity, so this is a reading, not a gate result.

Gates and tests (head a0a35cb30f)

• node scripts/pm/dispatch-gates.mjs --commands derived 127 commands from the 21-path diff. That is the same set before and after the merge.
• At 378d424915 (pre-merge) and again at a0a35cb30f, all 127 exited 0. --ran reports "127 derived, 127 run, 0 NOT-MEASURED, 0 UNRUN", with every exit code recorded.
• Also exit 0:
  • check:objectui-pin-citations --verify-anchors with objectui at the pin: 54 asserting citations match 2e818d0b5, and 7 content assertions are verified.
  • check:objectui-bump (20 assertions across 5 cases), check:sdui-lockstep, check-sdui-manifest, check:console-sha, check:console-injection, check:migration-registry.
  • @objectstack/spec check:generated: all 15 artifacts current after the merge.
• pnpm --filter @objectstack/spec test: 612 files, 18185 passed, 1 todo.
• pnpm --filter @objectstack/spec typecheck: exit 0.
• pnpm --filter @objectstack/sdui-parser test: 225 passed.
• eslint --no-inline-config on the 15 changed TS files: 0 errors and 0 warnings. The lint population is **/*.{ts,…} (eslint.config.mjs:971). The config enables no type-aware linting (:328), so this diff cannot move a judgment on an untouched file. Repo-wide pnpm lint is left to CI.

Acceptance notes

• ui/component.zod.ts's object-grid columns record (the ⚠️ paragraph on plugin-grid/src/useColumnSummary.ts:558-568) says, about ab1879721595, that the footer summary reads currency / defaultCurrency / precision / scale off an authored column. objectui#11588 (d768c3178, inside this pin) retires that read. The record is in the historical spelling and stays true of the pin it names, so it is left as is.
• Two anchors in records that cite no sha are outside the pin-citation gate's population and were not re-measured:
  • ObjectView.tsx:882, in ui/view.zod.ts's [ViewItemSchema 同时是授权形状和 Studio 往返的 wire 成员 —— 拆成两个 schema 还是保持宽松?(挡住 #4001 批 18 最后 2 站点) #5074] block and two view tests.
  • plugin-dashboard/src/index.tsx:204, in the ObjectMetricPropsSchema icon liveness test.
  • Both already read an unrelated line at ab1879721595, and they shift to :936 and :214 here.
• Writes: no PR assignee or label was written. This dispatch's write budget names neither.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

This PR changes 2 package(s): @objectstack/sdui-parser, @objectstack/spec, touching 14 documentable anchor(s). ⚠️ 2 changed file(s) yielded no anchor (packages/sdui-parser/objectui-lockstep.json, packages/spec/src/kernel/functional-completeness.ts), so the pages documenting them are NOT COVERED by this run — this is not a clean bill of health for those files.

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

• content/docs/data-modeling/analytics.mdx (via DatasetMeasureSchema (symbol, a top-level const))
• content/docs/protocol/objectui/layout-dsl.mdx (via ComponentPropsMap (symbol, a top-level const object))

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

• content/docs/releases/v15.mdx (via PageTabsProps (symbol, a top-level const object))
• content/docs/releases/v17/17-1.mdx (via ComponentPropsMap (symbol, a top-level const object))
• content/docs/releases/v17/17-3.mdx (via ComponentPropsMap (symbol, a top-level const object))
• content/docs/releases/v17/17-4.mdx (via ComponentPropsMap (symbol, a top-level const object))
• content/docs/releases/v17/17-5.mdx (via ComponentPropsMap (symbol, a top-level const object))

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

• 2 changed file(s) yielded no anchor (packages/sdui-parser/objectui-lockstep.json, packages/spec/src/kernel/functional-completeness.ts) — pages documenting those are invisible to this run
• 2 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 — 139 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 7e0066af7a8e05709dc60096c69bc85e299d0331 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from e9af031ca5f1edcc7f4e1a9c04b3f105731824a9 — the merge of head a0a35cb30f9451294889f99df9d931b9d4d66d3b into base 7e0066af7a8e05709dc60096c69bc85e299d0331, 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 e9af031ca5f1edcc7f4e1a9c04b3f105731824a9 && git checkout e9af031ca5f1edcc7f4e1a9c04b3f105731824a9
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 7e0066af7a8e05709dc60096c69bc85e299d0331 a0a35cb30f9451294889f99df9d931b9d4d66d3b && git checkout -B drift-repro 7e0066af7a8e05709dc60096c69bc85e299d0331 && git merge --no-ff a0a35cb30f9451294889f99df9d931b9d4d66d3b

node scripts/docs-audit/affected-docs.mjs --json 7e0066af7a8e05709dc60096c69bc85e299d0331

⚠️ 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 7e0066af7a8e05709dc60096c69bc85e299d0331 → pass the list as
args.docs, on the commit named under Which tree this was computed on.

[objectstack-fleet]

Contract review

Served-tier: CONTRACT_REVIEW_TIER
Head-sha: a0a35cb30f9451294889f99df9d931b9d4d66d3b
Local-runs: none

① Changesets — TRUE

• Breaking-commit list, measured myself on a full, non-shallow, non-partial clone of objectui: git log --format='%h %s' ab187972..2e818d0b5 | grep -E '^[0-9a-f]+ [a-z]+(\([^)]*\))?!:' yields exactly one commit, 83e3f8377 (objectui#11466). git rev-list --count --no-merges gives 22, as the PR says. Of the 18 .changeset/*.md files ADDED in the range, exactly two carry a breaking annotation, both added by 83e3f8377: 11466-envelope-object-metric-retired.md and 11466-node-slot-union.md. The 14 other changeset files 83e3f8377 touched are +2-line edits to files that already existed at ab187972 (already dispositioned on chore(objectui): bump the console pin to ab1879721595 (carries objectui f624f278, b0bf413c and ab187972) #21625), so they are correctly outside this bump's count.
• .changeset/console-2e818d0b51ec.md (@objectstack/console: minor) carries exactly one <!-- adr-0087: not-required (no-migration-prescription) ... --> marker that names both entries of 83e3f8377 and gives a reason for each. Every declared-breaking commit in the range is dispositioned.
• Claim (1) true: DeclaredNode is a @object-ui/types export (SchemaRenderer.tsx diff at the pin: schema: DeclaredNode | string | null | undefined); no packages/spec schema references it.
• Claim (2) true: packages/spec/src/ui/dashboard.zod.ts:1078-1081 declares DashboardWidgetSchema as strictObject({... guidanceSets: WIDGET_GUIDANCE_SETS }), and WIDGET_GUIDANCE_SETS (:184-190, set QUARANTINED_WIDGET_KEYS) names 'component' with the prescription "objectui-internal renderer capabilities, not part of the author-facing dashboard spec". A widget component key is refused by name, so the retired envelope form is not ObjectStack-authorable.
• .changeset/objectui-pin-citations-2e818d0b51ec.md (@objectstack/spec: patch, Clause-②: no): correct level. The only shipped text that moves is the FormField.span describe string and six migration-entry description strings (pin sha + corpus counts); no key, default, enum member or export moves (verified in ②), so patch is right and matches chore(objectui): bump the console pin to ab1879721595 (carries objectui f624f278, b0bf413c and ab187972) #21625's @objectstack/spec patch.
• Precedent form: both files are the same shape as chore(objectui): bump the console pin to ab1879721595 (carries objectui f624f278, b0bf413c and ab187972) #21625's console-ab1879721595.md / objectui-pin-citations-ab1879721595.md (script-derived entry list, "declared nowhere" section, single adr-0087 HTML-comment marker with the same opening paragraph, objectui range: trailer; spec patch with Clause-②: no). Like chore(objectui): bump the console pin to ab1879721595 (carries objectui f624f278, b0bf413c and ab187972) #21625's, the console changeset carries no Clause-② line of its own; check-adr-0087-registration.mjs:639 reads the arm from the PR body (readClause2Line(parsed.body)), and the PR body carries Clause-②: no.

② Review faces — TRUE, all citation/comment/describe-text only

• packages/spec/src/ui/component.zod.ts (+125/-39), action.zod.ts, dataset.zod.ts, view.zod.ts, kernel/functional-completeness.ts, six migrations/entries/semantic/18.*.ts and the regenerated migrations/registry.ts: every hunk is inside a /** ... */ or // comment, or inside a .describe(...) / description: string literal. No Zod builder, key, default, enum member, export or import changes (read the full diff).
• KEY CHECK — corpus counts are genuine, not a sparse-checkout artifact. On a full clone (git rev-parse --is-shallow-repository = false, no partialclonefilter): git ls-tree -r --name-only <sha> | wc -l = 10267 at ab187972 and 7579 at 2e818d0b5. The drop is .changeset/: 2728 tracked files at ab187972, 8 at 2e818d0b5 (objectui release b493919c7 consumed them). Lit-control spot-checks with git grep -o -F <term> <sha> | wc -l reproduce every recorded number at both pins: objectstack 16377 → 17227, @objectstack/spec 6665 → 7134, timeout 1348 → 1351, useState 2476 → 2477, TTL 180 → 182, tenant 1238 → 1317, Span 491 → 505, SpanSchema 56 → 57. Hit counts rise while file count falls because the release folded the changesets into the package CHANGELOGs. PASS.
• Anchor spot-checks at 2e818d0b5 (all hold): packages/plugin-dashboard/src/index.tsx:256 = 'object-metric', and :281 = { name: 'icon', type: 'string' }, (were :244 / :269 at ab187972, byte-identical text; file +14/-1); packages/types/src/objectql.ts:4720 = limit?: number; at both pins; content/docs/plugins/plugin-kanban.mdx still carries the limit rows (:174, :189); plugin-grid/src/ObjectGrid.tsx:5458 = keyboardNavigation: schema.keyboardNavigation ?? inlineEditable, and the file is byte-identical across the hop; core/src/actions/ActionRunner.ts:1497 = composeSuccessMessage(...), byte-identical; SchemaRenderer.tsx changes only its type import, docblock and schema prop type, as the record says. functional-completeness.ts cites ObjectCalendar.tsx / ObjectGantt.tsx / ObjectTimeline.tsx / ObjectMap.tsx / ListView.tsx as byte-identical; consistent with the measured 4-file change set of the hop.
• content/docs/references/ui/view.mdx: the two changed rows are the FormField.span row in the two generated tables that share FormFieldBaseSchema, and the cell text equals the new view.zod.ts:3319 describe string verbatim (ab1879721595 → 2e818d0b51ec, nothing else). Two mdx occurrences from one describe is the generator's shape, not a hand edit; the dev report says check:generated --fix wrote it and check:generated reports all 15 artifacts current.

③ Scope — TRUE, no ride-along, no release act

• 21 files: .objectui-sha, two changesets, scripts/sdui-manifest.record.json, packages/sdui-parser/objectui-lockstep.json, view.mdx, and 15 packages/spec/src files that are citation-only (three of them tests, comment-only as well). Nothing outside what the bump's gates require. sdui.manifest.json is not in the diff (sha256 0ead67c1... and 107 components unchanged in the record), consistent with no component input moving.
• Lockstep / manifest consistency: packages/sdui-parser/src at objectui is byte-identical across ab187972..2e818d0b5 (git diff --quiet), so the lockstep record moving only rev, revDate (2026-10-04T07:05:00+00:00 = the commit date of 2e818d0b5) and recordedAgainstPin, with blob/codes/predicate untouched, is correct. The manifest record's objectuiWorkspaceVersion 17.6.0 → 17.7.0 matches objectui package.json at the pin. objectui validate.ts:418 / :465 still grade type-mismatch / member-type-mismatch by arms.includes('enum'), so the finding(sdui-parser): the html tier's type-mismatch stays a warning for a string literal handed to an object-typed input — os build passes and the tile draws no number #21671 reading in the PR body is accurate.
• Pin target: 2e818d0b5 is objectui#11598, 973fc20f3 is its ancestor (card's floor satisfied). No objectui write, no tag, publish, Release or Version Packages action anywhere in the PR; version PR chore: version packages #21352 untouched.
• Head check-runs at review time: 14 completed/success (incl. Check Changeset, Governed Surface Queue Guard, Spec property liveness, Build Docs, Type Check · source gates), 1 skipped (opt-in tarball smoke), 17 still in_progress (Test Core 1-6, Build Core, Console Pin Gate, Lint & Repo Gates, Type Check · consumer/debt/workspace, Dogfood 1-3 + CLI, Temporal Conformance); 0 failures. Landing still waits on those going green.

Implemented-by: claude/issue-21696-objectui-pin-bump
Reviewed-by: session_01VDtqoecgES7ScQYGbFVDRv

VERDICT: PASS

---

Generated by Claude Code

[objectstack-fleet]

Console Pin Gate is red on this head (a0a35cb30f, job 111402467290), and the cause is not this PR's. Seat domain:devx#1, read at 2026-10-04T09:22Z.

• The failure: the step "Build the Console SPA at the pinned objectui SHA" prints ✗ Built console still carries the PUBLISHED @objectstack/spec, citing 1 of 38 published-only descriptions: "Extra vertical reference lines drawn like the Today marker ({ date, label?, color? })".
• Why it is not this PR's:
  • The same step fails at the old pin ab1879721595 on main's merge-queue builds since feat(spec)!: object-gantt markers, object-timeline mapping and both forms' fields take the shape each block reads; five objectui-held contracts reported as forks (#21464, S-objectui-held) #21699 (16d241a6af), for example run 37185775211.
  • The text the gate matches is objectui's own packages/plugin-gantt/src/index.tsx:190 registry description. That line is byte-identical at both pins.
  • feat(spec)!: object-gantt markers, object-timeline mapping and both forms' fields take the shape each block reads; five objectui-held contracts reported as forks (#21464, S-objectui-held) #21699's reword of the object-gantt markers describe turned that prefix into a "published-only" probe.
  • scripts/console-spec-probes.mjs chooseProbes tests it with a bare bundle.includes(...).
  • The dev reproduced this from a clean .cache and dist (exit 1, same text). The only bundle hit is in plugin-gantt-*.js. vendor-objectstack-*.js carries only this tree's wording.
• The fix is in flight elsewhere: card ci(console): the Console Pin Gate's staleness leg matches objectui's own gantt markers description as "the published spec", so every merge-queue build since #21699 is red #21709 (domain:spec, p1, dispatched) owns this gate defect. No fix PR exists yet. This PR does not widen into gate tooling.
  • A candidate approach the dev measured, for ci(console): the Console Pin Gate's staleness leg matches objectui's own gantt markers description as "the published spec", so every merge-queue build since #21699 is red #21709's owner to weigh: match a description only as the whole argument of .describe(<q>…<q>) for ", ' or backtick, in both legs of chooseProbes and in the mirrored replays in scripts/check-console-injection.mjs.
    • With the patched matcher, this dist passes: 26/26 injected-only descriptions present, 0/38 published-only present.
    • A real published 17.6.0 copy is still caught: 38/38.
• No re-run: the failure is deterministic.
• Next step: once ci(console): the Console Pin Gate's staleness leg matches objectui's own gantt markers description as "the published spec", so every merge-queue build since #21699 is red #21709 lands, this PR takes main in through a base merge, and the gate re-runs. Automerge stays enabled. The contract-review record on a0a35cb30f is re-recorded only if that merge changes reviewed content.

---

Generated by Claude Code

[github-actions]

⛔ merge queue 构建失败 — 先分诊，再决定要不要重排

队列构建 37192871034 红了。队列跑的是全量套件（PR 侧 CI 只跑 affected 子集），
所以失败的测试可能在本 PR 没碰过的包里 —— 那不是重排能修的。每次盲目重排都会让排在后面的所有 PR 重建一轮。

失败的 job（日志抽取，best effort）：

• Console Pin Gate — 失败步骤: Build the Console SPA at the pinned objectui SHA

  ✗ Built console still carries the PUBLISHED @objectstack/spec.
  

↳ 失败原因 是判读的关键：超时（Test timed out in … / Hook timed out in …）多半是负载/时序，不是本 PR 的回归；
断言（AssertionError: …）才指向真实的行为改变。两者的 FAIL 行长得一模一样，只有这一行能区分。

⚠️ 断言这一侧有一类例外，判据是断言在测什么，不是它是不是 AssertionError。 断言的对象是产品行为（一个值、一个形状、一次拒收）⇒ 照上面读：真实的行为改变，去查，⛔ 不要重排掉；
断言的对象是这次实验自身的有效性前提（跑完的耗时、负载下的先后、任何只在时间预算内才成立的条件）⇒ 它跟超时是同一类，同样对负载敏感，重排一次是合法的判别手段。
识别是机械的：断言的消息或它比较的值本身点名了一段时长、一个时间戳、一个耗时计数。实测过的一对 —— AssertionError: SecurityPlugin.init() ran: expected false to be true 测的是产品行为（真回归）；
AssertionError: this run took over a second, so second-precision stamps could have differed too: expected 1006 to be less than 1000 测的是实验前提：它守护的那条不变式当时是绿的，同一个 head 原样重排一次即成功。
穿着 AssertionError 外衣的时间测量，仍然是时间测量。（⛔ 这只改「怎么读一次红」，不改「哪些测试可以重排」——后者由别处管。）

跨 PR 相同签名（24h，按失败测试文件聚合）：

• ⚠️ 本次没有可用的聚合签名（日志里没有能解析出测试文件名的 FAIL 行）—— 这不是「没有同签名的其他 PR」，是这一轮没测到。跨 PR 聚合本次不可用，请手工比对其他 PR 的同类评论。
• ⚠️ 24h 评论账本没读完（超过 5 页仍未读到窗口尽头），所以上面的「不同 PR 数」是下界，不是全量。

历史信号：

• 本 PR 过去 24h 无队列失败记录（首次）。
• 过去 24h 队列共有 9 个失败构建（不含本次）。

分诊清单：

1. 失败测试在本 PR 改动的包里 → 真回归，修 PR。
2. 失败测试与本 PR 无关 → 看上面的「跨 PR 相同签名」；已有汇总 issue ⇒ flaky/环境问题实锤，去那张 issue 上谈，修好前重排只会再烧一轮全队列。
3. 两者都不是 → 可能与同组 PR 语义冲突；等前面的 PR 落地或失败出队后再重排一次即可，不要连续重排。

---

Generated by Claude Code · merge-queue-triage workflow (#4859)