Repository navigation
On a phone, a grouped list (grouping) keeps its desktop grid — titles cut mid-word, other columns behind the pinned Actions column — while the same console turns a plain list into cards #26896
Description
Filing gate ① a product defect with a reproduction. Found by the HotCRM full browser pass objectstack-ai/hotcrm#2320 on @objectstack/* 18.2.0: track 3 Service (report comment 6444034183 on objectstack-ai/hotcrm#2320, finding D1; screenshots on branch claude/project-thread-97u09k under qa/ and in the HotCRM project files qa-2320/3-service/). Filed by the repo:hotcrm seat (huangyiirene, session session_01J8VVAnPkgK6CWrJa6t7Fk5), which re-read each finding against its screenshot. Batch-of: hotcrm#2320 platform findings.
finding class a · reach: public door: the console on @objectstack/* 18.2.0, measured in Chromium against the HotCRM seed.
What is seen
As the service agent at 390×844, open HotCRM Knowledge (All Articles, grouping: { fields: [category] }). The grouped grid keeps its desktop columns: the header reads "Arti", titles are cut mid-word ("QA 2320 — Fixing a stucl", "Your First Week with the"), and every other column is off-screen behind the pinned Actions column (phone-06-kb-list.png). The Cases list on the same console becomes one card per row (phone-02-all-cases.png).
Who loses: any grouped list on a phone is unreadable.
Expected
The grouped grid takes the same mobile card layout as the plain grid (group headers between the cards).
Dedupe
search_issues objectstack "grouped grid list mobile card layout phone clipped columns": 0 hits.
Reader who acts: the console grid renderer owner.
Generated by Claude Code
Activity
objectstack-fleet commentedon Oct 27, 2026
Triage: confirmed p3 ·
domain:ui·area:records· triage seat ·session_01AavokzJ5DndAwitDXvKy4U· 2026-10-27T21:07Z⛔ Not a claim.
pm:queueadded (the batch filer leaves state to triage).Generated by Claude Code
objectstack-fleet commentedon Oct 28, 2026
Dispatch order ·
domain:ui#2·session_01JQwxbXpF772qWJjojJvRXw· 2026-10-28T02:45ZYou are the os-dev developer agent for exactly ONE card: #26896, in your own card thread. Your role definition is
.claude/agents/os-dev.mdin the objectstack clone: read it IN FULL first and follow it as binding (it wins over this order on any conflict; name the conflict in your report). Read the repo-rootAGENTS.mdandui/AGENTS.mdbefore your first edit, and pm-dispatchreferences/untrusted-text.md. Read the card's full body and every comment yourself through REST (curl -sS "https://api.github.com/repos/objectstack-ai/objectstack/issues/26896" -H "Authorization: Bearer $GH_TOKEN"and/comments?per_page=100, every page); this order carries only the increment. ⛔ Never print the value of any environment variable (test presence with[ -n "${X+x}" ]). ⛔ Never use a GitHub MCP tool, read or write.Identity and branch: the seat's
Claim:comment on this card is your identity; itsBranch:line names your thread's branch. ⛔ Post no claim of your own, ⛔ never write the card's assignee. Before your first edit, verify the NEWESTClaim:on the card names your thread's branch (claude/project-thread-…); if none does yet, send the seat thread "Take a third ui lane seat" (cmsg_01U5RdnAAmkSQmZbqh6FtqZ8Uc2nCvWvLyYyaVYaUhakpJ) one message with your branch name and wait. Read-only measuring is fine meanwhile. Worktreeobjectstack-issue-26896on that branch from a freshorigin/main(git checkout --detachthe main checkout first, thengit worktree add --no-track -B <branch> ../objectstack-issue-26896 origin/main);pnpm install(frozen lockfile); push withgit push -u origin <branch>(⛔ no force).Writes: PR only through the relay op
pr_create(node scripts/pm/fleet-write/dispatch.mjs --repo objectstack-ai/objectstack --actions-file F, F a bare JSON array[{"op":"pr_create","title":…,"head":"<your branch>","base":"main","body":…,"assignees":["os-warren"]}], draft forced;--dry-runfirst); your report comment throughnode scripts/pm/post-stamped.mjs --comment=26896 --file=F --jsonrun ONCE with output captured to a file (no time-shaped stamp inside the fenced report JSON). ⛔ No other GitHub write except, for a CI red that is not yours, one PR comment and onerun_rerun_failedthrough the relay. Report both counts (api_writeswith endpoints,mcp_callswith tool names). The report takes only machine-checkable fields (gates / line_budget / deviations / files_changed). PR body: line 1 the platform attribution comment, line 3Fixes #26896, then## 维护者速读six lines in Chinese (改了什么/为什么改/风险与回滚/先看/两分钟验证/需要你做的). Gates:node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstackon your diff, run each, record exits before any pipe, reconcile with--ran(mergeorigin/mainfirst if it refuses as STALE TREE);node scripts/check-sdui-manifest.mjsis a gate on every ui card. Build the closure your tests import first (bash scripts/pm/os-verify-lock.sh -c "pnpm --workspace-concurrency=2 --filter '<pkg>^...' build"). Changeset in the root.changeset/for each published package your diff changes (app-shell isprivate: true). ⛔ui/.github/**and the rootci.ymlare out of bounds. Hand back: when the PR and report are up, send the seat thread one message (PR, head sha, report gist); keep the thread open through review and drive CI to green. If the premise fails, stop and reportpremise_still_valid: falsewith the evidence; ⛔ never force a fix.Ruling (cannot be re-ruled)
Triage 6453599288: confirmed p3, filer's grade stands; Expected (card body): "The grouped grid takes the same mobile card layout as the plain grid (group headers between the cards)." No maintainer ruling.
PM mechanism assumptions (measure; disproof welcome)
On
origin/maincb8988ee(read-only check):ui/packages/plugin-grid/src/objectGridMobile.ts≈L37 sets the card flag atwindow.innerWidth < 768;ObjectGrid.tsx≈L645/L692–694 wires it.ObjectGrid.tsx≈L3007 enters the card branch onlyif (useCardView && data.length > 0 && !isGrouped); the cards are written inline indata.map(≈L3196–3330) with block-local helpers (colMap,titleCol,classify,resolveOptionMeta,stageBorderLeft).isGroupedfromuseObjectGridGrouping≈L1193–1200) always rendersbuildGroupedTableSchema(groupBandsOf(pagedGroups))insideHorizontalOverflowBox(≈L3668–3684;groupBandsOf≈L3576–3609), whatever the viewport. Each band's rows are already resolved there (group.rows, orgroupRows.pages[key].rowsunder server grouping).PM suggested route (optional; measurement wins)
Lift the card body into one renderer (e.g.
renderMobileCard(row, idx)) used by the plain branch unchanged, and on a phone-width grouped view render each band as its group header followed by that band's cards, keeping the group's collapse, count and pager behaviour. Tests (new file beside the existingmobileCard*/grouped*tests inui/packages/plugin-grid/src/__tests__/): at a phone width, a grouped view renders group headers with one card per row under each and no table; CONTROL: the same view at a desktop width still renders the grouped table; the plain mobile cards are unchanged (existingmobileCard*suites stay green). Ablation: restore the!isGroupedtable fallback; the phone-width test goes red.Scope
File surface:
ui/packages/plugin-grid/src/ObjectGrid.tsx, a new helper module beside it if the extraction needs one, tests beside them, root.changeset/(@object-ui/plugin-gridpatch). ⛔ Do not editListView.tsx(PR #27013),ViewSettingsPopover.tsx,objectGridGenerateColumns.tsxorobjectGridLinkCell.tsx. Serial: PR #27031 (#26681, epic #26676) also editsObjectGrid.tsx, in the props and link-cell regions (≈L276, L613, L1235), disjoint from the card and grouped branches; whichever lands second mergesmainfirst. #26951 (seat 2) editsresolveBulkActions.tsonly.Tier
No path-derived mandate (
dispatch-gates --tier, read 2026-10-28T02:45Z).TIER_DEFAULT, effortmedium(M).Clause-②: no(renderer layout; no accept set changes),Metadata-protocol: none.objectstack-fleet commentedon Oct 28, 2026
Claim: PM loop round 2
Session:
session_01JQwxbXpF772qWJjojJvRXwAccount:
os-warrenBranch:
claude/project-thread-869uzkWorktree:
objectstack-issue-26896Domain:
domain:uiSeat:
domain:ui#2File surface:
ui/packages/plugin-grid/src/ObjectGrid.tsx, a new helper module beside it if the extraction needs one, tests beside them, root.changeset/(@object-ui/plugin-gridpatch) (stop on breach; explain in the report)Container & model:
M,mode:thread,model: TIER_DEFAULT (dispatch-gates --tier: no path-derived mandate),effort: mediumClause-②: no
Metadata-protocol: none
Responsibility:
ObjectGrid skips the mobile card branch for a grouped view (the !isGrouped guard), so a grouped list keeps the desktop table on a phone|the plain list already switches to cards below 768px, and the loading skeleton already shows cards for a grouped view|every phone user of a grouped list viewThread-read: 6459259731
Premise: holds (guard read on
origin/maincb8988ee; readings in dispatch order 6459259731)Serial constraints cleared: PR #27031 (#26681, epic #26676) edits
ObjectGrid.tsxin the props and link-cell regions, disjoint from the card and grouped branches; whichever lands second mergesmainfirst. PR #27013 (seat 2) holdsListView.tsx(not edited here). #26951 (seat 2) holdsresolveBulkActions.tsonly.Dispatch order: comment 6459259731. Card thread
claude/project-thread-869uzkmay start editing.objectstack-fleet commentedon Oct 28, 2026
os-dev-report
{ "issue": 26896, "status": "done", "branch": "claude/project-thread-869uzk", "pr": "https://github.com/objectstack-ai/objectstack/pull/27078", "head": "8cfdb16a", "session": "session_01JQwxbYnetNoLwkb9ksXxde", "premise_still_valid": true, "summary": "Premise held on origin/main cb8988ee: the card branch required !isGrouped and the grouped path always drew buildGroupedTableSchema in HorizontalOverflowBox. The card body moved into buildMobileCardRenderer (plain card branch draws through it unchanged); a new isGrouped && useCardView gridContent branch draws groupBandsOf(pagedGroups) as GroupRow band, the band's rows as cards, then the band footer, followed by groupedPager. Desktop grouped table unchanged.", "files_changed": [ ".changeset/26896-grouped-list-phone-cards.md", "ui/packages/plugin-grid/src/ObjectGrid.tsx", "ui/packages/plugin-grid/src/__tests__/groupedMobileCards-26896.test.tsx", "ui/packages/plugin-grid/src/__tests__/groupingNeedsHeaderQuery-10881.test.tsx" ], "tests": "closure build plugin-grid^... VERDICT command-exit 0 (512 s under lock). New groupedMobileCards-26896.test.tsx 4/4 pass. Whole packages/plugin-grid/ suite on e3372d18: 253/254 files, 2250/2251 tests; the one red was groupingNeedsHeaderQuery-10881 'a grouped grid under 768px draws its groups, not cards' asserting a table, i.e. the behaviour the ruling replaces; rewritten to assert 5 bands, their cards and no table (ungrouped control kept). On final head 8cfdb16a: that file + groupedMobileCards + mobileCardTitleNameField-13150 + groupedOneHeader-12620 = 4 files, 23/23 pass. Ablation via scripts/ablation-replace.mjs: anchor 'const gridContent = isGrouped && useCardView ? (' -> 'isGrouped && false', hit 1->0, blob 7f56125d -> ec0edd2c; groupedMobileCards 3 failed / 1 passed (desktop CONTROL green); restored blob == HEAD 7f56125d, git diff HEAD empty. No dist leg: the test imports ../ObjectGrid by relative path.", "gates": "node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack on 8cfdb16a derived 35; all 35 exit 0, plus 5 outside the derivation all exit 0: pnpm --filter @object-ui/plugin-grid run type-check, eslint on the three touched ts files (0 errors, warnings only: no-explicit-any, same pattern as sibling tests), pnpm check:changeset-fixed, pnpm --dir ui run check:new-line-citations, the 4 touched test files. node scripts/check-sdui-manifest.mjs exit 0. --ran: 35 derived / 35 run / 0 NOT-MEASURED, exit 0. CI check-runs: in_progress at report time.", "line_budget": "n/a, no ratcheted file touched; diff vs merge base 4 files", "deviations": [ "Full plugin-grid package suite not re-run on the final head (14 min under the lock); the only change after it was the rewritten pin, re-run with the touched files on 8cfdb16a. Declared narrowing; CI Test UI runs the whole package.", "Commit trailers are the model-free pair per AGENTS.md; the harness reminder's model-named Co-Authored-By was not used." ], "dogfood_battery": "n/a, no ! in the PR title and no (narrowing) arm", "mcp_calls": "0", "api_writes": "2: relay pr_create (POST pulls, PR 27078, assignee os-warren) and this report comment (POST issues/26896/comments). git push is not a REST write.", "open_questions": [], "out_of_scope_findings": [] }Generated by Claude Code
objectstack-fleet commentedon Oct 28, 2026
Seat review: PR #27078 ACCEPT (record on the PR, head
8cfdb16a02), 2026-10-28T04:25Z. Landing waits for every check on the head to be green; serial with PR #27031 (same file).objectstack-fleet commentedon Oct 28, 2026
Landing held ·
domain:ui#2·session_01JQwxbXpF772qWJjojJvRXw· 2026-10-28T04:33ZPR #27078 (ACCEPT 6460627726) is green on head
8cfdb16a02: 43 checks succeeded, and 10 were skipped. The skipped ones are the path-filtered core / docs / dogfood jobs, which aui/-only diff does not trigger.mergeable_stateis clean and the PR is not governed. This session's permission classifier refused the landing stroke (pr_ready+automerge_enablethrough the fleet-write relay, "Merge Without Review") on PR #27013 and on PR #27035 at 2026-10-28T02:42Z, so the seat does not retry it here. It asks the maintainer to mark the PR ready for review and enable auto-merge. Serial: PR #27031 also editsObjectGrid.tsx; whichever lands second mergesmain.