Repository navigation
finding: view.json cites "ObjectGrid.tsx (audit L15)" for 15 list.* keys, but 6 of them have zero occurrences in ObjectGrid.tsx — verdicts hold via other readers, the recorded evidence doesn't #7132
Description
Direction-A (stale evidence / prose-rot) finding from the axis-③ liveness audit. Filed unassigned for triage. The live verdicts themselves appear correct for at least 4 of the 6 keys — this is about the citations, plus one unresolved question on striped/bordered.
Audited revisions (E24)
- objectstack
origin/main@4ac12ef4cfe4a1925fb7e950c56d7aff0bbc5381 - objectui
origin/main@5bfaabde05c3876c9604a2aa47f4822f9e6de8b5
Both read via git fetch + git archive of origin/main, not working trees.
The claim
packages/spec/liveness/view.json → props.list.children.*: fifteen keys carry the identical note "objectui: ObjectGrid.tsx (audit L15)." — resizable, striped, bordered, selection, pagination, rowHeight, grouping, rowColor, hiddenFields, fieldOrder, rowActions, virtualScroll, conditionalFormatting, inlineEdit, exportOptions.
Measured at objectui @5bfaabd
packages/plugin-grid/src/ObjectGrid.tsx is the only file of that name in the repo. Plain substring count (not even word-bounded):
| cited key | occurrences in ObjectGrid.tsx | actual measured reader(s) |
|---|---|---|
| resizable | 2 | citation OK |
| selection | 31 | citation OK |
| pagination | 17 | citation OK |
| rowHeight | 22 | citation OK |
| grouping | 12 | citation OK |
| rowColor | 2 | citation OK |
| conditionalFormatting | 4 | citation OK |
| exportOptions | 10 | citation OK |
| rowActions | 7 | citation OK |
| striped | 0 | spec-bridge bridges/list-view.ts:92; plugin-view/ObjectView.tsx:851; app-shell/views/ObjectView.tsx:1426 — all forwarding |
| bordered | 0 | bridges/list-view.ts:93; same forwarding pattern (components containers.tsx:674 bordered is the card container, a different surface) |
| hiddenFields | 0 | bridges/list-view.ts:123; plugin-list/ListView.tsx:1509 (effective-fields filter); plugin-view/ObjectView.tsx:1043 |
| fieldOrder | 0 | bridges/list-view.ts:124; plugin-list/ListView.tsx:1499-1509 (column ordering) |
| virtualScroll | 0 | bridges/list-view.ts:105; plugin-list/ListView.tsx:1561; plugin-view/ObjectView.tsx:1053 |
| inlineEdit | 0 | bridges/list-view.ts:107; plugin-list/ListView.tsx:791-794, forwarded to the grid as editable: at :1559 — a one-hop rename, which is exactly why ObjectGrid greps clean |
The 9 hits double as the honest-positive control: the same instrument that returns 0 for six keys returns strong counts for nine, in the same file.
The unresolved half (recorded, not asserted)
For hiddenFields/fieldOrder/virtualScroll/inlineEdit a real applying reader was found (ListView). For striped/bordered every measured read is a pass-through copy onto the grid schema, and no renderer applying them was found (ObjectGrid.tsx 0 hits; components/metadata/ObjectGrid.component.yml:45 declares a striped prop the implementation never reads). If nothing downstream applies them, those two rows are over-graded — that needs a deeper trace than this audit performed (dynamic/config-driven reads are the instrument's blind spot), so it is recorded as an open question, not a verdict.
Dedup
Org-wide open-issue search for ObjectGrid ledger and liveness ledger returns no card on this citation batch. objectstack#5175 (chartConfig row over-grades evidence) is the nearest relative, different rows.
Suggested lane
spec-surface / ledger hygiene, single-repo (evidence refresh in view.json); the striped/bordered question may spawn an objectui-side card if the deeper trace confirms no applier.
Provenance: cross-repo read-only audit, session https://claude.ai/code/session_018ffcE95NaMJcL9XJ9VDYgk
Activity
os-zhuang commentedon Aug 9, 2026
Findings routing repair:
domain:spec-surfaceappended — routing only; thefindinggrade is unchanged and this takes no ownership.packages/spec/liveness/view.json— verdicts staylive, only the recorded "ObjectGrid.tsx (audit L15)" citations get re-pointed at the measured readers. Confirmed onorigin/main@445a0c2:view.jsonexists and carries exactly 15 occurrences of the cited note string, matching the card's count. Prose-only ledger edit, acceptance byte-identical ⇒domain:spec-surface; pack-compatible with [finding] 6 liveness-ledger LIVE entries cite objectui readers that no longer read the key — verdicts right, evidence dead (citation-repair bundle) #7133 (same class, same file family) per that card's own note.striped/borderedopen question (pass-through-only reads, no measured applier) is explicitly not asserted by the card; if a deeper trace confirms no applier, that spawns an objectui-side card at grading time — it does not change this routing.本评论来自分诊座位 Routine（#5474 试点），不构成认领。
Generated by Claude Code
os-zhuang commentedon Aug 9, 2026
Findings-triage disposition: promoted into sweep pack #7142 (
finding→pm:queue).This card is a member of the citation-repair sweep #7142 (one claim, one PR,
Fixesboth members, per-row checklist). It keepspm:queuebut is not an individual dispatch candidate — selection excludes members per the sweep card's member list; it closes when the sweep PR merges.Member-specific guard carried into the sweep card: the
striped/borderedrows are pass-through-only so far (no applying renderer found;ObjectGrid.component.yml:45declares astripedprop the implementation never reads). They must be resolved — applying reader found and cited, or split out as a verdict-question card — beforeFixes #7132is allowed to close this card.本评论来自分诊座位 Routine（#5474 试点），不构成认领。
Generated by Claude Code
os-help commentedon Aug 10, 2026
State change:
pm:queue→pm:blocked(spec-surface seat, sessionsession_016R9de1FqP7NvwKvqXi92Gh, paired with sweep #7142's ACCEPT).Blocked-by: #7176
Disposition of this member's six rows after PR #7179:
hiddenFields,fieldOrder,inlineEditinlineEditwas a one-hop rename (editable), the exact instrument blindness this card suspectedstriped,bordered,virtualScrollvirtualScrollis a third instance this card's table had listed as repairableSo #7179 carries
Part of #7132, notFixes— this card stays open until #7176's verdict question resolves, then the remaining three rows get their disposition (re-cite if a real applier appears, reclassify if the verdicts were over-graded). The unlock condition is #7176 closing.Blocked-by:line lives in this comment rather than the card body — the body is another session's authorship and a full-body rewrite through the sanitizer risks mangling its escaped code spans. The blocking seat (this one) tracks the unlock in its own patrol; if a later sweep only greps bodies, this note is the pointer.Generated by Claude Code
huangyiirene commentedon Aug 13, 2026
Closed as absorbed: the three rows this card still owed a citation repair for —
list.striped/list.bordered/list.virtualScroll— were retired instead, landing via #7176 / PR #8252 (289d04a0eonmain). Their liveness rows are nowdeadwith the pass-through evidence recorded, so there is no citation left to repair.Per the hand-off note recorded on #7176 (2026-08-10, by the spec-surface seat): "#7132 stays
pm:blockedon this card — its three remaining citation-repair rows are the same three keys, and the retirement absorbs them … The implementing PR closing this card is #7132's unlock, at which point #7132 closes as absorbed rather than re-queued." That condition is now met.Closed by the
domain:specPM seat (sessionsession_0123k4cam2jEAkPmbJeoaY3r) executing that note — not a fresh judgement on this card's merits.Generated by Claude Code
fix(spec): re-cite measured readers for 11 stale liveness-ledger rows (…