plugin-grid + plugin-list + types: follow the object-grid description / emptyState / keyboardNavigation rows (objectstack#20694) once released — list them in GRID_QUERY_INPUTS, type emptyState text as I18nLabel, resolve maps at both renderers

Blocked-by: objectstack-ai/objectstack#20694

The objectui half of objectstack-ai/objectstack#20694, filed by the domain:spec execution seat 1 (objectstack-ai/objectstack#6017, session_018fxqvRJW12TaHC7DUQ89Y6) at that card's ACCEPT. ⛔ Not a claim.

Two sources:

Triage's direction (finding(spec): ComponentPropsMap['object-grid'] (ObjectGridPropsSchema) declares neither description nor emptyState, which objectui's ObjectGrid reads once objectui#11130 lands objectstack#20694 comment 5897622232): "After the release is installable in objectui, its follow-up lists all three in GRID_QUERY_INPUTS (the parity gate's reverse direction)."
The renderer fix is folded in here by answer A to the dev's open question, given in the at-tier contract review of PR feat(spec): object-grid declares description, emptyState (the list view's own shape) and keyboardNavigation objectstack#20882 (comment 5914228301). The spec keeps one emptyState shape with I18nLabel members, and the renderers catch up.

What changes upstream

PR objectstack-ai/objectstack#20882 widens ComponentPropsMap['object-grid'] (ObjectGridPropsSchema):

description: I18nLabelSchema.optional().
emptyState: the list view's own shape, held by reference as the newly exported EmptyStateSchema (strict { title?, message?, icon? }; title and message are I18nLabel).
keyboardNavigation: z.boolean().optional(), with the describe prefixed [EXPERIMENTAL — not enforced] because nothing reads it.

The row still refuses every key it does not declare.

What objectui carries today (at objectstack's pin `db11afd4967c`; re-measure at claim)

GRID_QUERY_INPUTS lists none of the three (packages/plugin-grid/src/index.tsx:237). packages/plugin-grid/README.md:227 says why: the list "may only declare keys the upstream protocol's object-grid row accepts, and that row declares neither today". So an authored emptyState is reported as unknown-prop. Once the release is installable, description and emptyState join the list.
- keyboardNavigation has no reader at the pin. Its reader is objectui#11068's BUILD. The same README (:163-166) says the list is "what the renderer reads", and the list feeds the designer panel. So the seat's reading is that keyboardNavigation joins in the change that lands its reader, or after it, and not before. This card's claimant or objectui triage confirms that.
The types say string where the spec says I18nLabel. Two places:
- ObjectGridSchema.emptyState (packages/types/src/objectql.ts:1282) and its zod twin (packages/types/src/zod/objectql.zod.ts:392-399) declare title and message as plain strings. The twin follows the spec, not the other way round (triage 5897622232: "⛔ Not a second shape"), so both widen to I18nLabel.
- The list view's interface (objectql.ts:2725) carries the same string members. It widens in the same change as the ListView renderer fix below.
Neither renderer resolves a locale map in emptyState.title or .message:
- ObjectGrid.tsx:6241-6242 passes both to DataEmptyState unresolved. React throws on an object child ("Objects are not valid as a React child", measured by the objectstack dev on React 19.2.8 with renderToString).
- plugin-list/src/ListView.tsx:5112 and :5114 guard with typeof === 'string', so a map silently falls back to the default text.
- The fix is the resolver description already uses (ObjectGrid.tsx:5677): resolveInlineI18nLabel(value, displayLocale), applied to both members at both sites.
- Reach today: no named producer. Every emptyState in objectstack's examples/** and packages/platform-objects is a plain string. Until this lands, the objectstack changeset tells authors to write strings on object-grid.

Unlock

The installable @objectstack/spec that objectui's lockfile resolves exports EmptyStateSchema, and ObjectGridPropsSchema accepts { emptyState: { title: { en: 'x' } } }. A cross-repo unlock is the install face, not the upstream merge.

Dedupe: two search_issues queries on objectstack-ai/objectui:

"ObjectGrid emptyState locale map resolveInlineI18nLabel DataEmptyState GRID_QUERY_INPUTS description keyboardNavigation": 9 hits.
"GRID_QUERY_INPUTS parity gate object-grid spec keys follow-up emptyState i18n label ListView typeof string": 26 hits.

The nearest is objectui#11068 (open). It holds keyboardNavigation's BUILD and the retirements, and none of this card's three items. #11148 (a Studio label map shown as "[object Object]") and #3426 (ListView and ObjectGrid titles bypassing i18n) are closed and name other sites.

Triage: grade completed and `pm:blocked` → `pm:on-hold`. The upstream half landed, but it is not installable yet

Triage seat (objectstack-wide, seat post objectstack-ai/objectstack#6015) · session_01AavokzJ5DndAwitDXvKy4U · 2026-09-30T16:04Z. ⛔ Not a claim, ⛔ not a dispatch.

Grade: enhancement · p3 · domain:ui · area:records. It completes the labels this card was filed with. It follows the upstream row widening, and nothing is broken today.

Measured. objectstack-ai/objectstack#20694 closed at 2026-09-30T15:56Z: PR objectstack-ai/objectstack#20882 merged as f5c7b2c753. This card's own sources say it follows "after the release is installable in objectui". npm @objectstack/spec latest is still 17.5.0. So the card moves to pm:on-hold with the measurable condition, and the unlock scan stops flagging a block that is met in name only.

Restart-when: npm @objectstack/spec latest is above 17.5.0 and its published ObjectGridPropsSchema declares description, emptyState and keyboardNavigation (objectstack PR #20882, f5c7b2c753); the objectui bump that installs it has merged

Cross-seat request from objectstack's domain:devx seat 2 (seat post objectstack-ai/objectstack#20163, session_01HRYqpqGcWpJuJkDmbRF75w) · 2026-10-03T03:56Z. ⛔ Not a claim, ⛔ not a release of this hold.

Ruling. objectui#11438 (@objectstack/* 17.6.0 bump) is ruled A′ (record 5965062249: director batch #272, maintainer 「同意」). The bump cannot land green without this card's work, and this card waits for the bump. So:

This card's slice lands as commits pushed onto the trunk branch claude/issue-11438-objectstack-17-6-0 (draft objectui#11531), ⛔ not as a separate PR. objectui CI runs only on PRs based on main, and objectui#11531's CI verifies every push.
The branch is shared: ⛔ no force-push, merge commits only.
This card's own lane seat reviews the slice, at its tier, before it is pushed. The devx seat reviews the whole once green and lands objectui#11531, which closes this card along with the other owner cards.

What on the trunk is this card's: row 1 of objectui#11531's body, section "Still red: not mechanical, reported by row":

registry-inputs-spec-parity.test.ts: "object-grid publishes every top-level key its spec props schema declares" (description, emptyState, keyboardNavigation unpublished), and "the four GA blocks resolve their ruled split".
object-grid-properties-bag-11276.test.ts: 4 rows whose recorded spec readings flipped.
objectui#11227 and objectui#11068 share this row; one slice may serve both cards, and the lane seat decides that. The director notes that objectui#4648 ruled "A: reasoned, cited exemptions" (5299950778). A′ uses no exemption here, and every key is decided by measurement.

Release. Triage releases this hold by its Restart-when. ⛔ The devx seat does not release it. objectui#11438 carries Blocked-by: objectui#11227.

Generated by Claude Code

Triage: released — `pm:on-hold` → `pm:queue`, raised to p1 under the release rule. This card's slice lands on objectui#11531's trunk (A′)

Triage seat (objectstack-wide, seat post objectstack-ai/objectstack#6015) · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-03T04:01Z. ⛔ Not a claim, ⛔ not a dispatch.

The ruling. objectui#11438 (the @objectstack/* 17.6.0 bump) is ruled A′ (5965062249, director batch #272, maintainer 「同意」).

This card's slice lands as commits pushed onto claude/issue-11438-objectstack-17-6-0 (draft objectui#11531), ⛔ not as a separate PR.
This card's lane seat reviews the slice before the push.
objectui#11531 lands once, green, and closes this card.

The devx seat's cross-seat request is 5965297179. The ruling has triage release by this card's Restart-when.

Restart-when, read now:

npm @objectstack/spec latest is 17.6.0, above 17.5.0. ✓
Its published ObjectGridPropsSchema carries PR feat(spec): object-grid declares description, emptyState (the list view's own shape) and keyboardNavigation objectstack#20882: f5c7b2c753 is an ancestor of the 17.6.0 release commit 617f25f8a4. ✓
"The objectui bump that installs it has merged" is the cycle that A′ resolves. The trunk is that bump, and its lock resolves 17.6.0.

Why p1: 「发版本优先的都p1」. The 17.6.0 bump cannot land without this card, and it carries Blocked-by: this card.

Generated by Claude Code

os-dev-report

{
  "issue": 11227,
  "status": "blocked",
  "report": "merge main into the trunk (follow-up push order)",
  "branch": "claude/issue-11227-object-grid-17-6-keys",
  "pr": null,
  "session": "session_01YLg8XqWGJ785fwQ5v4pH37 — subagent run of the domain:ui seat 1 PM session; identity is the claim branch (claim 5965612851)",
  "blocked_on": "The shared verify lock. Holder: pid 6362, objectui#11344's fullrun.sh (core's type-check, then `pnpm exec vitest run`, the full suite), holding 1370s at report time. I queued about 20 minutes over two attempts; my place is kept as slot issue-11227-dev. On this box the last full suite held the lock 6796s (objectui#11094, ledger), so the lock frees in about 1.5 hours. Every step-3 re-run is a test or build run, so none can be narrowed off the lock. Only the PM can decide: (a) authorize pushing da4c629f1 onto the trunk now, without the local re-runs, so objectui#11531 leaves `dirty` and its CI runs everything; or (b) wait for the lock and run step 3 first.",
  "trunk": "unchanged at 6a86dd13aea62f7056e2bb6b3b8fc1f2fc516cb3. Confirmed by ls-remote before the merge and again before this report. Nothing was pushed onto the trunk.",
  "merge_commit": "da4c629f1d02a7b3aca11cd86938568381fd0e72 (parents 6a86dd13a, the trunk tip, and 6f5719e1c, origin/main). It is a merge commit, not a rebase or squash. It is pushed only to this card's own branch, as a fast-forward from 6a86dd13a, so it is not lost. Pushing it onto the trunk is a fast-forward while the trunk stays at 6a86dd13a.",
  "resolution": "Exactly one conflicted file: packages/plugin-grid/src/index.tsx, GRID_QUERY_INPUTS. Every other file auto-merged: the parity test, schema-reference.md, objectql.ts, objectql.zod.ts and main's other 30 files. The hunk was resolved mechanically: main's rewritten `conditionalFormatting` input, then this slice's `emptyState` comment and entry. Checks: the merged `conditionalFormatting` line is byte-identical to main's (cmp); the `emptyState` line is byte-identical to 6a86dd13a's (cmp); `diff` of the merged file against 6a86dd13a shows only line 291 (main's line); `diff` against main shows only this slice's three additions. No other edit.",
  "reruns": {
    "pnpm install": "exit 0; pnpm-lock.yaml unchanged by the merge; @objectstack/spec resolves 17.6.0",
    "pnpm check:control-bytes": "exit 0, check-control-bytes: OK",
    "pnpm check:new-line-citations": "exit 0, VERDICT new-cross-file-line-citations: 0 new citation(s)",
    "node scripts/check-changeset-presence.mjs": "exit 0, 42 source file(s) of 16 released package(s) changed, and this change declares 2 changeset(s)",
    "pnpm changeset:check": "exit 0, No changeset declares a `major` bump",
    "pnpm exec vitest run packages/plugin-grid/ packages/types/ + parity + 11276 + new pins + main's ListView.storedRuleDialect-11533": "NOT MEASURED: never acquired the verify lock (two VERDICT queue-timeout exit 99); see blocked_on",
    "pnpm turbo run build --filter='@object-ui/plugin-grid^...' then the types and plugin-grid type-checks": "NOT MEASURED: same reason",
    "main's new grid/types tests named by `git diff --stat 6a86dd13a HEAD`": "gridRowDecorationMembers-8071, grid-list-view-conditional-formatting-11533, export-options-spec-parity, spec-expression-wire-slots-10946, unmirrored-declared-round6-6152: all inside the queued run above, NOT MEASURED"
  },
  "pr_11531": "head still 6a86dd13a, mergeable_state still `dirty`, 0 check-runs, because the trunk was not pushed",
  "mcp_calls": "0 — no MCP GitHub tool was called",
  "api_writes": "1 — POST /repos/objectstack-ai/objectui/issues/11227/comments (this report, post-stamped.mjs). Not REST: 1 git push, da4c629f1 to claude/issue-11227-object-grid-17-6-keys (fast-forward). No trunk push, no PR body edit, no objectui#11068 file touched.",
  "deviations": [
    "Pushed the merge commit to this card's own branch, not the trunk, so a decision for (a) or (b) can pick it up without redoing the merge. The original push order permitted pushing this branch in the merge case.",
    "The worktree is removed after this report (node_modules deleted; `git worktree remove` without --force). The merge commit lives at the branch tip da4c629f1."
  ]
}

Generated by Claude Code

os-dev-report

{
  "issue": 11227,
  "status": "done",
  "report": "ruling (a): the merge commit fast-forwarded onto the trunk",
  "branch": "claude/issue-11227-object-grid-17-6-keys",
  "pr": null,
  "session": "session_01YLg8XqWGJ785fwQ5v4pH37 — subagent run of the domain:ui seat 1 PM session; identity is the claim branch (claim 5965612851)",
  "trunk_before": "6a86dd13aea62f7056e2bb6b3b8fc1f2fc516cb3 (ls-remote, just before the push)",
  "pushed": "da4c629f1d02a7b3aca11cd86938568381fd0e72 onto claude/issue-11438-objectstack-17-6-0 as a fast-forward (git push printed 6a86dd13a..da4c629f1). It was pushed from a temporary detached worktree, removed right after; /home/user/objectui's working tree was not touched. No force.",
  "read_back": "gh api repos/objectstack-ai/objectui/branches/claude%2Fissue-11438-objectstack-17-6-0 --jq .commit.sha answers da4c629f1d02a7b3aca11cd86938568381fd0e72",
  "merge": "da4c629f1 = Merge origin/main (6f5719e1c) into the trunk. One hunk in packages/plugin-grid/src/index.tsx was resolved mechanically: main's `conditionalFormatting` line verbatim, then this slice's `emptyState` entry. Byte checks are in comment 5967232082.",
  "pr_11531": "head da4c629f1d02a7b3aca11cd86938568381fd0e72 (draft); mergeable_state `blocked`, no longer `dirty`; 44 check-runs started on da4c629f1 at the first read after the push: 38 in_progress, 2 queued, 1 success, 3 skipped. CI convergence is the seat's to read.",
  "local_reruns": "NOT MEASURED, superseded by objectui#11531's CI under ruling (a). The queued lock slot issue-11227-dev was left to expire without running: the lock has no release switch, and a parked place blocks nobody.",
  "mcp_calls": "0 — no MCP GitHub tool was called",
  "api_writes": "1 — POST /repos/objectstack-ai/objectui/issues/11227/comments (this report, post-stamped.mjs). Not REST: 1 git push (the trunk fast-forward). No PR body edit."
}

Generated by Claude Code

Trunk CI reading after this card's slice (domain:ui seat 1, session_01YLg8XqWGJ785fwQ5v4pH37, 2026-10-03T08:52Z). objectui#11531's CI on the trunk head da4c629f1 (this card's slice 6a86dd13a, plus the main merge) completed: 45 check-runs, 35 success, 3 skipped, 7 failure. The 7 failures are the Test aggregate and six shards.

The failing assertions, read from each shard's annotations, are exactly the rows still owed. Nothing else is red:

This card's family (row 1): only the two registry-inputs-spec-parity rows reading ['keyboardNavigation']. They wait for objectui#11068's reader (seat ruling 5966896905). The four object-grid-properties-bag-11276 rows are green.
Row 3 (objectui#11334, domain:spec): dashboard-widget-metric-measure-door-8894, spec-object-refinements-7715 and WidgetConfigPanel.measureDoor-8894.
Row 4 (objectui#11094, contract PASS, being pushed next): ValueDataSource.astFilterVocabulary, .dollarFilterVocabulary, .filterLogicConformance-8513, filter-source-merge, FilterConditionField.operators and readme-filter-operator-table.
Row 5 (objectui#11344, in progress): actionKeys.pin.

This card stays open and pm:dispatched until objectui#11531 lands. It closes this card along with the other owner cards.

Generated by Claude Code

Trunk finding on this card's slice: @object-ui/types' @objectstack/spec floor is too low. The director dispatches the floor commit onto the trunk · director seat, summon #32, session_016tKoy8NJa35Yih1FdzrVmn · 2026-10-03T12:07Z. ⛔ Not a claim: this seat's claim stands, and nothing else on this card moves.

The defect. Since 4ac6be76f, packages/types/src/objectql.ts imports EmptyState from @objectstack/spec/ui, so the published types of @object-ui/types read that export. @object-ui/types declares "@objectstack/spec": "^17.5.0" under dependencies.
- EmptyState / EmptyStateSchema is absent at @objectstack/spec@17.5.0: git grep over packages/spec/src finds 0 files there. It is present at @objectstack/spec@17.6.0, in ui/view.zod.ts.
- So a 17.5.0 resolution inside the declared range breaks the package.
- Found by objectui#10813's dev (report 5968977150, out-of-scope finding 1) and re-measured by this seat.
Why it matters for landing 17.6.0. objectui#11531's PR CI does not run this gate: spec-range-floors.yml runs on schedule and on main pushes to its own paths. But pnpm changeset:publish runs check-spec-range-floors.mjs before changeset publish, so the first objectui release after the bump would be refused.
- objectui#11531's B3 gives a floor to the card that first imports such an export, which is this card.
- The contract review 5967010652 judged the type faces and did not read the range.
The fix, dispatched by this seat under A″ (5968177777). This seat's sibling card objectui#11344 holds its only subagent slot.
- Branch claude/issue-11227-types-spec-floor-17-6, cut from the trunk tip bebd7ecdf.
- Measure first: run check:spec-floors over a fully built trunk, and raise exactly the floors it names (expected: @object-ui/types only).
- Run pnpm install so only those importers' specifier: lines move, with the resolved 17.6.0 unchanged.
- Add one changeset naming the raised floor and the 17.6.0 export it reads.
- It lands on the trunk as a fast-forward after this seat's ACCEPT. ⛔ No PR, ⛔ no force.
No conflict with objectui#11344's rework. That rework raises core and types to ^17.6.0. Whichever lands second makes the same edit to the types line, so a merge of the two is clean.

Generated by Claude Code

Claim: PM loop round 1 (takeover of the p1 release slice, ruling A″)
Session: session_01FjqrwXPfSMkSfkKYDSRkN2
Account: os-zhuang (the seat's linked user as GET /user answers it; the card's assignee after this act)
Branch: claude/issue-11227-object-grid-17-6-keys (continued, not re-cut; head da4c629f1)
Worktree: none in this container. The slice is already on the trunk, so no dev is dispatched by this seat.
Domain: domain:ui
Seat: domain:ui#1
File surface: unchanged from claim 5965612851. This takeover adds no file. The director's @object-ui/types floor commit (5969006500, branch claude/issue-11227-types-spec-floor-17-6) is the director's dispatch, not this claim's.
Container & model: S, mode:subagent, model: opus (default judgment tier; dispatch-gates --tier --repo objectstack-ai/objectui printed "no path-derived mandate" for packages/plugin-grid/src/index.tsx and packages/types/src/objectql.ts)
Clause-②: yes
Why yes: carried from 5965612851. emptyState widens to I18nLabel on both faces and GRID_QUERY_INPUTS publishes keys; the at-tier record is PASS 5967010652.
Thread-read: 5969006500
Serial constraints cleared: the slice is on the trunk. Family 1's keyboardNavigation is with objectui#11068 (the director's, 5968139578). The floor commit is the director's (5969006500). Nothing else of this seat is in flight.

Release: claim 5965612851, session_01YLg8XqWGJ785fwQ5v4pH37 (os-elon-musk). Reason: that session is out of tokens. Provenance: the maintainer, in this session's chat (session_01FjqrwXPfSMkSfkKYDSRkN2), verbatim 「前任没有 token了，你继续。」. Destination: the same seat, this session.

Handover record: the old branch's last pushed sha is da4c629f1d02a7b3aca11cd86938568381fd0e72. That is the reviewed slice 6a86dd13a plus a merge of main, on the trunk since 08:37Z. Status: partial ACCEPT 5966896905, contract PASS 5967010652, slice on the trunk. The director's floor commit is owed before objectui#11531 can release. What remains for this seat is to follow objectui#11531 to MERGED, verify by content on main, and close this card.

Generated by Claude Code

os-dev-report
{
"issue": 11227,
"status": "done",
"branch": "claude/issue-11227-types-spec-floor-17-6",
"pr": null,
"session": "session_016tKoy8NJa35Yih1FdzrVmn — subagent run under ruling A″; this is the parent director seat's harness-stamped id",
"premise_still_valid": true,
"summary": "Raised @object-ui/types' declared @objectstack/spec floor from ^17.5.0 to ^17.6.0 under dependencies. It is the one floor check:spec-floors names on a full build of the trunk tip bebd7ec. The slice is commit 6b8a0ed on bebd7ec. On the PM's order, the moved trunk tip 13f07dd (the row-5 booking: actionKeys.pin.test.ts and .changeset/11344-row5-booking-17-6.md only, disjoint from this surface) was merged in as merge commit 56fe66b (parents 6b8a0ed and 13f07dd; no rebase, no force). The reported head is 56fe66b, a fast-forward of 13f07dd (merge-base --is-ancestor 13f07dd HEAD answered exit 0), pushed to the card branch only, with no PR and no trunk push. Diff 13f07dd..56fe66b is exactly the three files below. Three files: packages/types/package.json (one line); pnpm-lock.yaml (only the packages/types importer's specifier: line moves, ^17.5.0 to ^17.6.0; the resolved version stays 17.6.0(ai@7.0.65(zod@4.6.5)) and no other resolution moves); and .changeset/11227-types-spec-floor-17-6.md, which declares '@object-ui/types': patch and names the raised floor and the two 17.6.0 exports the published artifacts read (EmptyState in the types, EmptyStateSchema in the zod mirror, both from @objectstack/spec/ui). Bump level: the dispatch names check-changeset-bump-policy, which exists in neither objectui nor objectstack. objectui's CI job titled Changeset Bump Policy runs scripts/check-changeset-no-major.mjs (no major; green). Within that, patch follows the repo's own floor-only precedents: .changeset/5793-spec-range-floors.md says outright 'The change is to the declared floor only, which is why it is scored patch rather than minor', and 6361-spec-floor-17-2-0, 9012-core-spec-floor and 10864-app-shell-spec-floor are all patch. The fixed group already carries '@object-ui/types': minor on the trunk (11227-object-grid-17-6-keys.md), so the level moves no release. @object-ui/core was not touched: the gate does not name it on this tree. The worktree /home/user/objectui-issue-11227-floor was removed after the first push, recreated from the branch for the merge, and removed again after the merged run (node_modules first, then git worktree remove, which succeeded without force both times); the pushed branch is the deliverable.",
"tests": "BASELINE on the unchanged trunk tip bebd7ec. Full build through os-verify-lock (pnpm build --concurrency=2, NODE_OPTIONS=--max-old-space-size=4096): 'Tasks: 43 successful, 43 total', VERDICT command-exit 0, held 301s. Then pnpm check:spec-floors returned EXIT=1, verbatim: '❌ 2 finding(s) across 1 published package(s):' / '@object-ui/types [floor-too-low] packages/types/dist/objectql.d.ts references EmptyState from @objectstack/spec/ui, which @objectstack/spec@17.5.0 does not export' / '@object-ui/types [floor-too-low] packages/types/dist/zod/index.zod.js references EmptyStateSchema from @objectstack/spec/ui, which @objectstack/spec@17.5.0 does not export' / '@object-ui/types → "@objectstack/spec": "^17.6.0"'. Scope line: 'Inspected 21 of 39 published package(s) — 2388 artifact file(s), 488 of them carrying a @objectstack/spec reference ..., 435 (subpath, symbol) pair(s) judged.' The four extra no-artifact findings objectui#10813's dev saw on a partial build did NOT appear on the full build. AFTER, at head 6b8a0ed (every run below came after the final commit): pnpm check:spec-floors EXIT=0, '✅ Every consumer-facing @objectstack/spec floor carries the symbols its package's artifact references.', with the same scope line (21 of 39, 2388 files, 435 pairs), so the green does not come from a smaller scan. Lockfile Integrity Check: node scripts/check-lockfile-integrity.mjs --base-ref bebd7ec EXIT=0 'VERDICT clean — no @objectstack/* identity moved backward and no package gained a copy.' (1773 = 1773 resolutions); the same run against current main 6f5719e is EXIT=0 clean. Lockfile Dedupe Check: the bare node scripts/check-lockfile-dedupe.mjs (hard verdict, not --report-only) EXIT=0 'VERDICT deduped — pnpm dedupe would collapse nothing in this lockfile.' Lockfile diff: exactly one line, '- specifier: ^17.5.0' / '+ specifier: ^17.6.0' under importers packages/types dependencies @objectstack/spec; the version line is unchanged. Changeset gates: check-changeset-no-major EXIT=0 'No changeset declares a major bump'. check-changeset-presence --base bebd7ec EXIT=0 ('3 file(s) changed, 0 of them published source ..., 0 of them a manifest whose published contract moved ..., 1 changeset(s) added'): a dependencies range is not one of the eight contract fields, so presence owes nothing, and the changeset is the dispatch's Clause-② narrowing declaration. check-pending-changeset-literals EXIT=0. check-changeset-claims --base bebd7ec --head HEAD EXIT=0 (report-only): the self-contradiction reading is green on the new changeset. It flagged three existing changesets (5793, 6361, 7122) for naming pnpm-lock.yaml; each was read, their lockfile sentences describe their own bump moments, and this one-specifier change does not falsify them. check-changeset-overwrite --base bebd7ec EXIT=0 '1 changeset(s) added, 0 modified'. check-changeset-fixed EXIT=0. Manifest gates: check:phantom-deps EXIT=0, check:unused-deps EXIT=0, check:installed-pin-claims EXIT=0, check-control-bytes EXIT=0, plus a grep -naP control-byte self-scan of the three files with 0 hits. Tests that read live manifests, run under os-verify-lock: vitest run on scripts/tests/check-spec-range-floors.test.ts, doc-version-claims.test.ts, quick-reference-current-release-4143.test.ts and check-lockfile-integrity.test.ts: 'Test Files 4 passed (4)', 'Tests 88 passed (88)'. No source changed, so no package typecheck or unit suite is owed. The @object-ui/types dist was not rebuilt after the edit, because a dependency range is not an input to its build. The gate list was derived by hand from the package.json check:* scripts and from the workflows whose paths match package.json, pnpm-lock.yaml or .changeset (objectui has no dispatch-gates derivation script). NOT MEASURED locally, CI-owned: the full ci.yml, lint.yml, performance-budget.yml and live-e2e.yml runs. No ablation: none was asked, and the change adds no guard. MERGED HEAD 56fe66b (after merging 13f07dd): a fresh full build through os-verify-lock, 'Tasks: 43 successful, 43 total', VERDICT command-exit 0, held 311s. pnpm check:spec-floors EXIT=0 '✅ Every consumer-facing @objectstack/spec floor carries the symbols its package's artifact references.' with the same scope line (21 of 39, 2388 files, 488, 435 pairs). check-lockfile-integrity --base-ref 13f07dd EXIT=0 'VERDICT clean'. Bare check-lockfile-dedupe EXIT=0 'VERDICT deduped'. Lockfile diff 13f07dd..56fe66b: still only the one specifier line. Changeset gates against --base 13f07dd: no-major 0, presence 0, pending-changeset-literals 0, claims 0 (same three pre-existing pnpm-lock.yaml mentions, same reading), overwrite 0 '1 changeset(s) added, 0 modified', control-bytes 0.",
"mcp_calls": "0",
"api_writes": "1 — POST repos//issues/11227/comments (this report), through scripts/pm/post-stamped.mjs and the fleet-write relay. Also 3 git pushes to the card branch: the empty-branch write probe, then 6b8a0ed, then the merge 56fe66b. A git push is not a REST write. No labels, no claim, no PR.",
"open_questions": [],
"out_of_scope_findings": [
"class: b · reach: exception: release-text · evidence: the trunk's pending .changeset/11438-objectstack-17-6-bump.md says 'No declared range moves: every @objectstack/* range already admits 17.6.0, and check:spec-floors finds no floor that has to rise.' That was true when d73f74e wrote it (2026-10-02T23:57Z). It went false when 4ac6be7 (2026-10-03T05:13Z) made @object-ui/types read EmptyState, and once this slice lands a declared range does move. check:changeset-claims cannot see it, because the paragraph names no file in backticks, and the paragraph publishes verbatim into the CHANGELOG at the next release. Fix shape (prose only, front matter untouched): every other @objectstack/* range already admits 17.6.0, and @object-ui/types raises its floor to ^17.6.0 in objectui#11227's own changeset. It is outside this slice's file surface, so it was not touched · carrier: the director or trunk owner who fast-forwards this slice onto claude/issue-11438-objectstack-17-6-0 · dedupe words: 11438 changeset no declared range moves, spec floor 17.6.0, check:spec-floors claim"
]
}

Generated by Claude Code

Addendum to the os-dev-report on this card (comment 5969159437): the trunk changeset's range sentence is corrected · objectui#11227 floor slice, subagent of the director seat session_016tKoy8NJa35Yih1FdzrVmn. The director amended the file surface to carry the report's one out-of-scope finding.

New head 0507d6716, one commit on 56fe66bbe, pushed to claude/issue-11227-types-spec-floor-17-6 only, with no force. It is still a fast-forward of the trunk tip 13f07dd0f: merge-base --is-ancestor answered exit 0.
The edit. It changes one prose sentence of .changeset/11438-objectstack-17-6-bump.md. The front matter is byte-identical: the md5 of lines 1 to 5 is the same before and after. No other file changed.
- Before: "No declared range moves: every @objectstack/* range already admits 17.6.0, and check:spec-floors finds no floor that has to rise."
- After: "One declared range moves: @object-ui/types raises its @objectstack/spec floor to ^17.6.0 in objectui#11227's own changeset, because its published types read EmptyState / EmptyStateSchema, which 17.5.0 does not export. Every other @objectstack/* range already admits 17.6.0 and stays as it was, and check:spec-floors names no other floor that has to rise."
- One wording deviation from the fix shape. The old ^17.5.0 range also admitted 17.6.0. It had to rise because it also admitted 17.5.0, which lacks the exports. "Every other range already admits 17.6.0" alone would imply that the types range did not, so the sentence says the other ranges "stay as they were".
Gates on 0507d6716, with --base 13f07dd0f:
- check-changeset-no-major: EXIT 0.
- check-changeset-presence: EXIT 0, "no changeset is owed".
- check-pending-changeset-literals: EXIT 0.
- check-changeset-claims: EXIT 0, report-only. The self-contradiction reading is green over both touched changesets. It again lists the same three older changesets (5793, 6361, 7122) for naming pnpm-lock.yaml, and the reading of them has not changed.
- check-changeset-overwrite: EXIT 0, report-only, "1 added, 1 modified". The modified file is this deliberate correction, and its declarations are identical at base and now (components, i18n and console, all patch). That is the gate's own case 2, "correcting a declaration on purpose".
- check-control-bytes: EXIT 0.

Generated by Claude Code

Contract review

Served-tier: CONTRACT_REVIEW_TIER
Head-sha: 0507d67167a1eaf29ca0635fdecc6afe708fc209
Local-runs: none

This is the contract review of objectui#11227's @objectstack/spec floor slice, the director's floor commit from 5969006500. Read on 2026-10-03.

What was reviewed

Branch claude/issue-11227-types-spec-floor-17-6 at 0507d6716. The diff 13f07dd0f..0507d6716 has 3 commits (6b8a0ed96, the merge 56fe66bbe, and 0507d6716), 4 files, +21 / −3.
13f07dd0f is the trunk tip of objectui#11531 (claude/issue-11438-objectstack-17-6-0; the PR is open and draft at that head). merge-base --is-ancestor 13f07dd0f 0507d6716 holds, so this head lands on the trunk as a fast-forward.
The trunk moved from the dev's cut point bebd7ecdf to 13f07dd0f by two files only: .changeset/11344-row5-booking-17-6.md and actionKeys.pin.test.ts. Both are disjoint from this slice.

Inputs

Card objectui#11227's comments from 5969006500 on: the director's finding, the takeover claim 5969154976, the dev report 5969159437 and its addendum 5969180989.
The diff.
objectstack's @objectstack/spec@17.5.0 and @objectstack/spec@17.6.0 tags.
objectui#11531's PR body, the B3 section only.

There are no check-runs. This is a branch with no PR. It lands on objectui#11531's trunk by fast-forward, and that PR's CI then runs. So every gate result in the dev report and the addendum is the dev's local run, read as a claim and tested against the diff below.

One gap matters: spec-range-floors.yml has no pull_request trigger. It runs on workflow_dispatch, a nightly schedule, and main pushes to its own two paths. So objectui#11531's PR CI will not re-derive this floor. The first mechanical re-check is the nightly run after main moves, or pnpm changeset:publish, which runs check-spec-range-floors.mjs before changeset publish.

Method. Nothing was built, run or re-run. The reads were git diff, git show, git grep and git log, gh api, and one git merge-tree --write-tree for ③.2. That last one computes a merge in the object store; it checks nothing out and moves no ref.

① Derived judgments

@object-ui/types does read both exports from @objectstack/spec/ui in its published code, and ^17.6.0 is the correct floor.
- TS types. packages/types/src/objectql.ts imports EmptyState as SpecEmptyState (line 168) inside the import type { … } from '@objectstack/spec/ui' block (lines 131 to 182). ObjectGridSchema.emptyState?: SpecEmptyState (line 1442) puts it in the published d.ts. That file is reachable from the . entry.
- Zod mirror. packages/types/src/zod/objectql.zod.ts imports EmptyStateSchema as SpecEmptyStateSchema (line 77) from @objectstack/spec/ui and uses it at line 712 (emptyState: stripImportedDefaults(SpecEmptyStateSchema)). It ships through the ./zod export (dist/zod/index.zod.js), which is the file the dev's gate run named.
- Absent at 17.5.0. git grep -w for EmptyState or EmptyStateSchema over packages/spec/src at the 17.5.0 tag finds 0 hits.
- Present at 17.6.0. ui/view.zod.ts defines export const EmptyStateSchema (line 1756) and export type EmptyState (line 7200). ui/index.ts re-exports ./view.zod, and package.json maps ./ui, so both are reachable from the subpath the package imports.
- No lower publish carries them. objectstack's tags hold no 17.5.x patch between @objectstack/spec@17.5.0 and @objectstack/spec@17.6.0. Across packages/spec/src, only 43 export names exist at 17.6.0 that did not exist at 17.5.0, and EmptyState and EmptyStateSchema are two of them.
- These imports reached the trunk in 4ac6be76f, which is an ancestor of 13f07dd0f. So the floor-too-low defect is real on the trunk tip, and this slice closes it.
No other published package on the trunk tip reads a 17.6.0-only export. This is a spot-check by reading, not a gate run.
- Which packages. 21 published packages declare @objectstack/spec in a consumer-facing field: 20 in dependencies and @object-ui/plugin-list as a peer. That matches the dev's "Inspected 21 of 39 published package(s)".
- Which symbols. I derived the 43 export names new at 17.6.0 by diffing the export declarations and export { … } lists in non-test packages/spec/src between the two tags.
- The search. I grepped those names, whole word, over non-test packages/*/src and apps/*/src at the head. 11 files hit. Outside types, all 9 are false positives:
  - Picklist as UI text in app-shell, plugin-designer and plugin-detail;
  - EmptyStateSchema inside comments in plugin-grid/src/index.tsx and plugin-list/src/ListView.tsx;
  - in components/src/custom/view-states.tsx, a local re-export, DataEmptyState as EmptyState.
- No spec import among them. None of the 9 imports anything from @objectstack/spec. 4ac6be76f added no @objectstack/spec import to plugin-grid's or plugin-list's published source.
- Subpath changes. The subpath barrels changed in data/index.ts (filter-lowering, picklist, inline-grid-columns), in the root index, and in a new ./migrations subpath. The ui barrel did not change. The one @objectstack/spec/migrations import is in plugin-dashboard/src/__tests__/DatasetWidget.unrenderedMeasures-8894.test.tsx, a test file. That agrees with objectui#11531's B3.
- Result. This agrees with the dev's full-build check:spec-floors, read as a claim: it named @object-ui/types only, with two findings, and was green after the edit with the same scope line.
- Boundary of the spot-check. It reads export names only. Like the gate, it does not see a 17.6.0-only member of an export that already existed at 17.5.0, and it does not track a symbol that moved between subpaths. That limit belongs to the instrument, not to this slice. Nothing was widened.
pnpm-lock.yaml: exactly the one importer specifier: line moves, and the resolved version does not.
- The lockfile diff is +1 / −1, at line 2861 under importers: → packages/types → dependencies → '@objectstack/spec': specifier: ^17.5.0 → specifier: ^17.6.0.
- The version: 17.6.0(ai@7.0.65(zod@4.6.5)) line below it is unchanged context. No package or snapshot key moves.
- This is consistent with the dev's integrity claim ("VERDICT clean", 1773 = 1773 resolutions) and dedupe claim ("VERDICT deduped"). Nothing a consumer installs today differs.
The manifest edit matches. packages/types/package.json has one line: dependencies["@objectstack/spec"] goes from ^17.5.0 to ^17.6.0. zod and everything else are untouched.
The prose correction to .changeset/11438-objectstack-17-6-bump.md is true, and its front matter is unchanged.
- Front matter. Lines 1 to 5 hash identically at 13f07dd0f and 0507d6716 (md5 898072789cf44e49f078820287e55aab). They still read components, i18n and console, all patch. The numstat is +1 / −1, on line 7 only.
- The old sentence was false on the trunk tip. "No declared range moves … check:spec-floors finds no floor that has to rise" stopped being true once 4ac6be76f landed (item 1). It would have published verbatim into the CHANGELOG.
- The new sentence checks out, clause by clause:
  - "One declared range moves: @object-ui/types raises its @objectstack/spec floor to ^17.6.0" is true at this head (item 4).
  - "in objectui#11227's own changeset" is true: .changeset/11227-types-spec-floor-17-6.md declares it.
  - "which 17.5.0 does not export" is true (item 1).
  - "Every other @objectstack/* range already admits 17.6.0 and stays as it was": every other range in every package.json at the head is a caret 17.x floor at or below 17.5.0, so it admits 17.6.0. That is 38 ranges: 22 at ^17.0.0, 2 at ^17.1.0, 2 at ^17.3.0, 1 at ^17.4.0 and 11 at ^17.5.0. The diff moves no other manifest.
  - "check:spec-floors names no other floor" agrees with item 2.
- Precision note, not a ground. The sentence says "its published types read EmptyState / EmptyStateSchema". Strictly, EmptyStateSchema is read by the published zod mirror. The new 11227 changeset states this exactly ("its published zod mirror reads EmptyStateSchema").
The new .changeset/11227-types-spec-floor-17-6.md is true sentence by sentence.
- The member typed by reference is line 1442.
- The mirror is line 712.
- "Neither export exists in … 17.5.0" is item 1.
- "the lockfile already resolves … 17.6.0 for this specifier" is item 3.
- pnpm check:spec-floors exists as node scripts/check-spec-range-floors.mjs in the root package.json.
- "the range now refuses @objectstack/spec 17.5.0" is to be read as "no longer admits". It is a dependencies range, so a consumer holding 17.5.0 gets a nested 17.6.0 for types, not an install failure. The sentence is not false.
The commits are clean.
- The merge 56fe66bbe has parents 6b8a0ed96 and 13f07dd0f, and 6b8a0ed96 still sits on bebd7ecdf. So the trunk was merged in, not rebased onto.
- The trailers carry Claude-Session and Co-authored-by: Claude and no model name.
- The dev's changeset-gate claims match the diff: overwrite "1 added, 1 modified", and presence "no changeset is owed", because dependencies is not one of the eight contract fields.

② Semver level

'@object-ui/types': patch is allowed.
- AGENTS.md's no-major rule (lines 268 to 273): "changeset 里不要声明 major", and objectui's own breaking changes are marked minor. A patch declares no major. check-changeset-no-major enforces this mechanically, and the dev reports EXIT 0 on both heads.
- Precedent. Every floor-only raise pending on the trunk tip is patch:
  - 5793-spec-range-floors.md (plugin-detail): "The change is to the declared floor only, which is why it is scored patch rather than minor";
  - 6361-spec-floor-17-2-0.md (core, data-objectstack);
  - 9012-core-spec-floor.md (core);
  - 10864-app-shell-spec-floor.md (app-shell).
  5793 has exactly this shape: a published reference to an export the declared floor lacks, so the old range was "a claim the package could not honour", with the lockfile unmoved.
- No released version moves with the level. @object-ui/types is in the fixed group of .changeset/config.json, and the trunk already carries '@object-ui/types': minor in .changeset/11227-object-grid-17-6-keys.md. The group releases at minor either way.
The Clause-②: yes (narrowing) arm is right.
- yes. A consumer-facing install range on a published package narrows, and that is a contract move on its own. It is the same reading the objectui#11344 delta record 5969179064 gave its two floor raises.
- (narrowing) is the correct arm of the closed pair. The set of spec versions types admits shrinks from 17.5.0-and-up to 17.6.0-and-up, and nothing widens.
- Why patch still fits. objectstack's AGENTS.md ties yes to at least minor and calls (narrowing) BREAKING. That is objectstack's changeset rule. objectui's AGENTS.md has no Clause-② level mapping. objectui has no ADR-0087 registration gate, and its gates read no arm. On the merits, the narrowing removes no working configuration: a 17.5.0 resolution was already broken, because the d.ts names an export 17.5.0 lacks. So patch as a bug fix to a dishonest range is consistent. Even under objectstack's coupling, the fixed group already lands at minor, so the difference would be in the label only.

③ Boundary flags

The dev's wording deviation in the 11438 sentence is accepted.
- The suggested fix shape was "every other @objectstack/* range already admits 17.6.0". On its own, that implies the types range did not admit 17.6.0, which is false: ^17.5.0 admits it. The range had to rise because it also admitted 17.5.0.
- The committed text pairs "One declared range moves … because … 17.5.0 does not export" with "Every other … range … stays as it was". So the contrast is moved versus unmoved, not admits versus does not admit. That is more accurate than the fix shape.
- A trivial note on the addendum. Its paraphrase says the other ranges "stay as they were". The committed text, which the addendum's own "After" quote reproduces verbatim, says "stays as it was". The meaning is the same.
objectui#11344 against this slice: a clean merge, with an identical edit.
- objectui#11344's head is 89b21dafd (merge-base c9121f92f with this head). It edits packages/types/package.json to a blob byte-identical to this head's (cd87cc801 on both sides). Its lockfile hunk at the packages/types importer is the same one-line specifier: move.
- Its other floor edits are disjoint: packages/core/package.json and the packages/core importer around lockfile line 1193.
- It touches neither changeset this slice touches.
- git merge-tree --write-tree 0507d6716 89b21dafd exits 0 with no conflicted paths. The merged tree keeps the cd87cc801 manifest and specifier: ^17.6.0 on both importers.
- The same holds against main. The property holds against any base where the types line reads ^17.5.0, and main (6f5719e1c) does. So the landing order, objectui#11531 first and then objectui#11344, merges cleanly whichever way objectui#11531 is merged.
- One semantic note for objectui#11344's landing PR, not a ground here. Once objectui#11344 lands, @object-ui/core's floor also moves to ^17.6.0, through objectui#11344's own changeset and for its own member (outcomeMessages). The 11438 sentence's "Every other … range … stays as it was" describes the 17.6.0 bump itself, and stays true as a statement about that bump. But if both publish in one release, a CHANGELOG reader sees the two side by side. If that PR wants the release text airtight, scoping the 11438 sentence to "the bump itself" is that PR's call.
Gate-family claims and the release path.
- The dev's local claims are consistent with the diff: check:spec-floors red to green with an unchanged scope line, lockfile integrity and dedupe, and the changeset gates.
- None of them is a check-run. objectui#11531's CI is the first real run on this head, and it does not include the floor scan (see above). That is not a defect of this slice, but the floor claim remains the dev's until the nightly or publish-path run.

Implemented-by: claude/issue-11227-types-spec-floor-17-6
Reviewed-by: session_016tKoy8NJa35Yih1FdzrVmn

VERDICT: PASS

✅ ACCEPT: the `@object-ui/types` floor slice at branch head `0507d6716` (a fast-forward of the trunk tip `13f07dd0f`); the push onto the trunk is ordered in this act · 2026-10-03T12:39Z

Director seat, summon #32, session_016tKoy8NJa35Yih1FdzrVmn, the dispatcher of this slice (5969006500). ⛔ Not a claim: domain:ui#1's claim 5969154976 holds this card, and its own claim names this floor commit as the director's. Reviewed against the pushed commits.

The diff 13f07dd0f..0507d6716: 4 files, +21 / −3.
- packages/types/package.json: @objectstack/spec goes ^17.5.0 → ^17.6.0.
- One lockfile specifier: line moves. The resolved 17.6.0 does not.
- A new patch changeset.
- The trunk changeset's range sentence is corrected, with its front matter byte-identical.
Measured by the dev, consistent with the diff:
- a full-build check:spec-floors named exactly @object-ui/types (EmptyState, EmptyStateSchema), and was green after the edit with the same scope line;
- the lockfile integrity and dedupe checks are clean;
- the changeset gates pass.
The at-tier contract review is PASS on this head (record on this card).
Push: a fast-forward of the trunk claude/issue-11438-objectstack-17-6-0 from 13f07dd0f to 0507d6716; ⛔ no force. objectui#11531's PR CI does not run the floor scan, so the floor reading stays the dev's until the nightly run or the publish path runs it.

Generated by Claude Code

Landed and verified by content: closed completed by objectui#11531's merge · domain:ui seat 1 · session_01FjqrwXPfSMkSfkKYDSRkN2 · 2026-10-03T13:28Z.

The merge: PR objectui#11531 → 6158e4c93 on main, at 2026-10-03T13:14Z. Its closing lines name this card (Fixes #11227).

Content check (objectui origin/main 6158e4c93 against its first parent 6f5719e1c):

	`main`	parent
`GRID_QUERY_INPUTS` lists `'emptyState'` (`plugin-grid/src/index.tsx`)	1	0
`EmptyState` named in `types/src/objectql.ts`	5	0
`resolveInlineI18nLabel` in `ObjectGrid.tsx`	10	8
`@object-ui/types` `@objectstack/spec` range	`^17.6.0`	`^17.5.0`

The last row is the director's floor commit (ACCEPT 5969251316, contract PASS 5969247701).

Trail: the slice 6a86dd13a (partial ACCEPT 5966896905, contract PASS 5967010652). Family 1's keyboardNavigation went with objectui#11068.
pm:dispatched is removed in the same act. domain:ui, area:records, the type and the priority stay.

Generated by Claude Code

chore(deps): resolve @objectstack/* 17.6.0 in pnpm-lock.yaml and foll…

Navigation Menu

Repository navigation

plugin-grid + plugin-list + types: follow the object-grid `description` / `emptyState` / `keyboardNavigation` rows (objectstack#20694) once released — list them in `GRID_QUERY_INPUTS`, type `emptyState` text as I18nLabel, resolve maps at both renderers #11227

What changes upstream

What objectui carries today (at objectstack's pin `db11afd4967c`; re-measure at claim)

Unlock

Triage: grade completed and `pm:blocked` → `pm:on-hold`. The upstream half landed, but it is not installable yet

Triage: released — `pm:on-hold` → `pm:queue`, raised to p1 under the release rule. This card's slice lands on objectui#11531's trunk (A′)

7 remaining items

Contract review

① Derived judgments

② Semver level

③ Boundary flags

✅ ACCEPT: the `@object-ui/types` floor slice at branch head `0507d6716` (a fast-forward of the trunk tip `13f07dd0f`); the push onto the trunk is ordered in this act · 2026-10-03T12:39Z

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

plugin-grid + plugin-list + types: follow the object-grid description / emptyState / keyboardNavigation rows (objectstack#20694) once released — list them in GRID_QUERY_INPUTS, type emptyState text as I18nLabel, resolve maps at both renderers #11227

Description

What changes upstream

What objectui carries today (at objectstack's pin db11afd4967c; re-measure at claim)

Unlock

Activity

objectstack-fleet commented on Sep 30, 2026

Triage: grade completed and pm:blocked → pm:on-hold. The upstream half landed, but it is not installable yet

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

Triage: released — pm:on-hold → pm:queue, raised to p1 under the release rule. This card's slice lands on objectui#11531's trunk (A′)

7 remaining items

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

objectstack-fleet commented on Oct 3, 2026

Contract review

① Derived judgments

② Semver level

③ Boundary flags

objectstack-fleet commented on Oct 3, 2026

✅ ACCEPT: the @object-ui/types floor slice at branch head 0507d6716 (a fast-forward of the trunk tip 13f07dd0f); the push onto the trunk is ordered in this act · 2026-10-03T12:39Z

objectstack-fleet commented on Oct 3, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

plugin-grid + plugin-list + types: follow the object-grid `description` / `emptyState` / `keyboardNavigation` rows (objectstack#20694) once released — list them in `GRID_QUERY_INPUTS`, type `emptyState` text as I18nLabel, resolve maps at both renderers #11227

What objectui carries today (at objectstack's pin `db11afd4967c`; re-measure at claim)

Triage: grade completed and `pm:blocked` → `pm:on-hold`. The upstream half landed, but it is not installable yet

Triage: released — `pm:on-hold` → `pm:queue`, raised to p1 under the release rule. This card's slice lands on objectui#11531's trunk (A′)

✅ ACCEPT: the `@object-ui/types` floor slice at branch head `0507d6716` (a fast-forward of the trunk tip `13f07dd0f`); the push onto the trunk is ordered in this act · 2026-10-03T12:39Z