List user-filter chips draw the raw field key ("category") until the object definition arrives: UserFilters resolves the label from objectDef, which ListView holds at null until getObjectSchema answers #12890

Activity

1. objectstack-fleet

   added

   finding

   on Oct 17, 2026
2. objectstack-fleet commented on Oct 17, 2026

   

   objectstack-fleetBot

   on Oct 17, 2026

   ContributorAuthor

   More actions

   Triage: p2 · area:i18n · domain:ui · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-17T14:56Z

   Triage seat (objectstack-wide, seat post objectstack-ai/objectstack#6015). ⛔ Not a claim, ⛔ not a dispatch.

   • Fresh read on objectui origin/main d2b21a1225: the defect holds.
     • packages/plugin-list/src/ListView.tsx:1727 holds objectDef at null, and :2301 resets it.
     • UserFilters.tsx still falls back to the machine key at four draw sites: :762, :788, :845 and :1282 (f.label || f.field).
     • objectDefLoaded (ListView.tsx:1728) already says whether the definition is pending.
   • Grade p2. This is the objectui half of the maintainer's 2026-10-16 walk (cloud#2915 item 2, p2). 我的软件包 shows category before 分类.
   • Scope:
     • While the definition is pending, a chip with no authored label draws a placeholder (or nothing) at all four sites, ⛔ never the key.
     • Once the definition arrives, the chip draws the field's translated label.
     • An authored label is unchanged.
     • A test beside UserFilters.tsx pins the pending render.
   • Lane: domain:ui (packages/plugin-list).

   ---

   Generated by Claude Code
3. objectstack-fleet

   added

   area:i18nThe customer's own language, across UI, metadata and notificationsThe customer's own language, across UI, metadata and notifications

   domain:uiobjectui ui stream: fix lands on the published library or apps — objectui execution seatobjectui ui stream: fix lands on the published library or apps — objectui execution seat

   pm:queue

   priority:p2

   pm:dispatched

   and removed

   pm:queue

   on Oct 17, 2026
4. objectstack-fleet

   assigned

   baozhoutao

   on Oct 17, 2026
5. objectstack-fleet commented on Oct 17, 2026

   

   objectstack-fleetBot

   on Oct 17, 2026

   Last edited by objectstack-fleet

   ContributorAuthor

   More actions

   Claim: PM loop round 3 (priority:p2, triage scope 6235382361)
   Session: session_01XVp19E1MYUTFM3dnkwz8gB
   Account: baozhoutao (the seat's linked user as GET /user answers it; the card's assignee)
   Branch: claude/issue-12890-userfilter-pending-label
   Worktree: objectui-issue-12890
   Domain: domain:ui
   Seat: domain:ui#2
   File surface: packages/plugin-list/src/UserFilters.tsx (all four f.label || f.field draw sites: while the object definition is pending, a chip with no authored label draws a placeholder or nothing, never the key; once it arrives, the field's translated label; an authored label unchanged); packages/plugin-list/src/ListView.tsx only to hand objectDefLoaded (or equivalent) down, if needed; tests beside UserFilters.tsx; one .changeset/12890-*.md (@object-ui/plugin-list: minor; amended in place from patch, see Clause-②). Stop on breach; explain in the report.
   Container & model: S, mode:subagent, model: opus (default tier; dispatch-gates --tier --repo objectstack-ai/objectui reads no path-derived mandate)
   Clause-②: yes
   Clause-② reason: amended in place after review of PR #12905. The measured fix adds an optional objectDefPending member to UserFiltersProps, which @object-ui/plugin-list's entry exports (export type { UserFiltersProps }), so a published prop surface widens. It owes one CONTRACT_REVIEW_TIER review, and the changeset levels minor.
   Responsibility: objectui's own user-filter chip, which falls back to the machine key | the list already tracks whether the object definition is loaded (objectDefLoaded) | every viewer of a list with unlabelled user filters, on each load (cloud 我的软件包)
   Thread-read: 6235382361
   Serial constraints cleared: no open objectui PR touches UserFilters.tsx or ListView.tsx, and no pm:dispatched claim names them (read at claim). objectui#12784 (the date-range chip) is a different site in the same file family; it is not in flight. This seat's objectui#12889 and objectui#12892 are disjoint.
   Claimed at: 2026-10-17T16:07Z
6. objectstack-fleet commented on Oct 17, 2026

   

   objectstack-fleetBot

   on Oct 17, 2026

   ContributorAuthor

   More actions

   os-dev-report

   {
     "issue": 12890,
     "status": "done",
     "branch": "claude/issue-12890-userfilter-pending-label",
     "pr": "https://github.com/objectstack-ai/objectui/pull/12905",
     "session": "session_01XVp19E1MYUTFM3dnkwz8gB",
     "premise_still_valid": true,
     "summary": "Premise held on origin/main c999d4c017. Four `f.label || f.field` sites in UserFilters.tsx drew the key: the dropdown chip (also used for the More overflow), LookupValuePicker field.label, the text placeholder and the toggle button. ListView's objectDefLoaded was never passed down (A1: a new optional prop was needed). ListView now passes objectDefPending={!objectDefLoaded}; UserFiltersProps gains an optional objectDefPending (default false). While it is pending, those sites draw the label if there is one, otherwise a Skeleton (chip and toggle) or an empty string (placeholder and picker label). resolveFields/useSafeFieldLabel are unchanged. Once settled the key fallback is unchanged: a field the settled definition does not name, or a failed getObjectSchema (settles with objectDef null), still draws the key, as does a host that omits the prop.",
     "tests": "All at head 85588ab6ce. New UserFilters.pendingLabel-12890.test.tsx (5 tests: PENDING all dropdown sites incl. overflow/placeholder/picker; PENDING toggle; CONTROL authored label while pending; SETTLED key as before; ARRIVAL via ListView with zh resources and a gated getObjectSchema: skeleton, then 分类). Ablation via objectstack scripts/ablation-replace.mjs in wrap mode: restored the chip site to {f.label || f.field}. Anchor x1 -> x0; blob ee06df6d9966 -> ad54466e63dc. Result: 'Tests 2 failed | 3 passed (5)'; PENDING and ARRIVAL failed with 'expected category not to contain category'. Restore proved: blob after == HEAD == ee06df6d9966, git diff HEAD empty. No build is needed for this ablation: the tests import ../UserFilters by relative path, i.e. source. pnpm exec vitest run --maxWorkers=1 packages/plugin-list/ under the lock: Test Files 150 passed (150), Tests 1523 passed (1523), VERDICT command-exit 0. Closure build turbo run build --filter=@object-ui/plugin-list^... exit 0; plugin-list type-check (tsc --noEmit && tsc -p tsconfig.test.json) exit 0, and --listFiles counts the new test (1). eslint narrowed to the 3 changed source files (root flat config, --format json): 3 files, 0 errors. Pre-existing warnings only; the diff adds 0 any. Type-aware linting is not enabled, so the verdict on untouched files cannot move. Battery all exit 0: phantom-deps self-import unreferenced-sources doc-example-readers handler-key-reads side-effects-array element-data-source-declaration spec-symbols action-forward-parity designer-field-key-parity icon-record-names i18n-keys i18n-drift i18n-designer-parity control-bytes new-line-citations, check-hand-rolled-comment-mask, check-changeset-fixed, check-changeset-presence ('3 source file(s) of 1 released package(s) changed, and this change declares 1 changeset(s)'). check:spec-floors exit 1 'no-artifact' (workspace not fully built): NOT MEASURED, expected per dispatch. CI: in_progress, not awaited.",
     "mcp_calls": "0",
     "api_writes": "3 so far — git push of the branch (empty probe and then the commit; git, not REST); 1 relay pr_create (POST /repos/objectstack-ai/objectui/pulls draft + POST /issues/12905/assignees, fleet-write request fw-20261017T163624Z-22b85c, read-back identical 4313/4313 bytes, assignees baozhoutao); 1 relay comment (POST /repos/objectstack-ai/objectui/issues/12890/comments, this report). Labels: zero writes (objectui, not named by the dispatch).",
     "open_questions": [],
     "out_of_scope_findings": [
       "carrier: none (承接者:无) · noted in PR Acceptance notes, not filed · the deprecated toggle mode never receives objectDef, so after the definition settles an unlabelled toggle chip still draws the key (before and after this PR); stored toggle configs only, no producer measured · dedupe words: toggle user filter label objectDef, ToggleFilters field key, deprecated toggle chip label"
     ],
     "gates": {"closure_build": 0, "type_check_plugin_list": 0, "vitest_new_file": 0, "vitest_plugin_list_dir": 0, "eslint_changed_files": 0, "battery_16_named_plus_new_line_citations": 0, "check_hand_rolled_comment_mask": 0, "check_changeset_fixed": 0, "check_changeset_presence": 0, "check_spec_floors": "1 (no-artifact; NOT MEASURED)"},
     "files_changed": ["packages/plugin-list/src/UserFilters.tsx", "packages/plugin-list/src/ListView.tsx", "packages/plugin-list/src/__tests__/UserFilters.pendingLabel-12890.test.tsx", ".changeset/12890-userfilter-pending-label.md"],
     "deviations": ["The test lives in packages/plugin-list/src/__tests__/, where every existing UserFilters.*.test.tsx lives, read as 'beside UserFilters.tsx'.", "Worktree removed after the PR opened."]
   }

   ---

   Generated by Claude Code
7. objectstack-fleet commented on Oct 17, 2026

   

   objectstack-fleetBot

   on Oct 17, 2026

   ContributorAuthor

   More actions

   Review: ACCEPT — objectui PR #12905 · domain:ui seat 2 · session_01XVp19E1MYUTFM3dnkwz8gB · 2026-10-17T17:24Z

   Reviewed on GitHub at head 5d0f34e, against the card's Wanted, the triage scope (6235382361) and the claim (6236336823, amended in place: Clause-②: yes and minor, after this seat's review found the new prop on an exported type).

   • Shape: the body says Fixes #12890; the second line is Clause-②: yes. The changeset is @object-ui/plugin-list: minor (patch round 5d0f34e; it was patch). The body's Published surface section names the one widening.

   • The change (packages/plugin-list):

     • Pending, not the key. ListView hands its objectDefLoaded state down as objectDefPending. While it is set, every site that named a chip f.label || f.field draws the label if there is one and otherwise a skeleton or no text:
       • the dropdown chip, also used for the More overflow;
       • the person-picker's label;
       • the text-search placeholder;
       • the deprecated toggle button.
     • One helper, chipName, is the single rule; resolveFields and useSafeFieldLabel are unchanged.
     • Settled, as before. Once the definition settles, a field the definition does not name, or a failed getObjectSchema, still falls back to the key. A host that omits the prop keeps the previous behaviour.
     • Published surface: UserFiltersProps (exported from the package entry) gains the optional objectDefPending?: boolean, default false. That is a widening, so Clause-②: yes.

   • Pins: UserFilters.pendingLabel-12890 (5 tests):

     • PENDING: every dropdown site, including overflow, placeholder and picker; PENDING toggle;
     • CONTROL: an authored label is drawn while pending;
     • SETTLED: the key as before;
     • ARRIVAL: through the real ListView with zh resources and a gated getObjectSchema, a skeleton, then 分类.

     Ablation: restoring the chip site's f.label || f.field turns PENDING and ARRIVAL red; the restore was proven by blob hash. packages/plugin-list (150 files, 1523 tests) is green; type-check exits 0; the changeset gates exit 0 on 5d0f34e.

   • Out-of-scope finding: the deprecated toggle mode is never handed objectDef, so after the definition settles an unlabelled toggle chip still draws the key, before and after this PR. Acceptance notes, not filed: only stored toggle configs reach it, and no producer was measured.

   • Process note: the dev declined the PR-body PATCH, per its role file (a dev writes the body once, at opening); this seat wrote the replacement body it prepared.

   • Trailers: both commits carry only the model-free pair.

   • Contract review: PASS at this head, CONTRACT_REVIEW_TIER, by an isolated reviewer and adopted by this seat (PR comment 6236904013). Its two non-blocking notes: while pending, an unlabelled chip button has no accessible name (the skeleton is aria-hidden), which is "nothing" under the card's scope; and a host with inline data shows the skeleton for one first commit before the effect settles.

   • CI at head: 33 success, 3 skipped, 0 failures (filter=latest, 36 runs), including Bundle Analysis (performance budget PASS), after one re-run of the per-shard Test gate (it read the shards while shard 8 was still running; stand-down 6237517319). Not governed (path audit).

   Next: pr_ready + automerge_enable.

   ---

   Generated by Claude Code
8. objectstack-fleet

   closed this as completedin #12905on Oct 17, 2026
9. objectstack-fleet commented on Oct 17, 2026

   

   objectstack-fleetBot

   on Oct 17, 2026

   ContributorAuthor

   More actions

   Landed · domain:ui seat 2 · session_01XVp19E1MYUTFM3dnkwz8gB · 2026-10-17T17:40Z

   PR #12905 merged through the merge queue as 1b88645, which is on main; Fixes #12890 closed this card. Verified by content: all 4 of the PR's files on main are blob-identical to the accepted head 5d0f34e. The merge commit's message names no model.

   What landed:

   • While a list's object definition is loading, an unlabelled user-filter chip draws a skeleton (or no text) instead of the raw field key, at every site that drew f.label || f.field. Once the definition arrives it draws the field's translated label; an authored label is drawn throughout.
   • UserFiltersProps gains the optional objectDefPending (default false; omitted = previous behaviour). The contract review (CONTRACT_REVIEW_TIER) PASSed (PR comment 6236904013).
   • Released as @object-ui/plugin-list: minor.

   For objectstack-ai/cloud#2915 (item 2, 我的软件包's category chip): cloud sees this once its .objectui-sha covers the merge commit above. This seat now removes pm:dispatched.

   ---

   Generated by Claude Code
10. objectstack-fleet

    removed

    pm:dispatched

    on Oct 17, 2026
11. akarma-synetal

    added a commit that references this issue on Oct 28, 2026

    fix(plugin-list): an unlabelled user-filter chip draws a skeleton, no…

    ...

    This issue will close once commit 1b88645 is merged into the 'main' branch.

    1b88645