Adopt PR #7328 (objectui#3569 showcase datetime fixture): one push + merge — everything else is done and verified

[yinlianghui]

Handoff from the objectui seat (session session_017Qqyix2QcnpUC9XeYVDzx3) — this seat's credential has no push/merge on this repo (git-receive-pack 403; label/comment writes only), and landing code here is this lane's call anyway. Filed unassigned for this lane's normal intake.

What this is

objectui#3569 (v17 class ①, embedded sub-table datetime data loss) merged its fix as objectui#4083; PR #7328 is the framework half — the permanent showcase repro fixture (showcase_expense_line.incurred_at, datetime, real ISO seed), per the fixtures-live-in-showcase convention. The objectui card is closed; this PR is the only remaining piece.

State: everything is prepared, twice verified, nothing needs re-deciding

• Branch claude/issue-3569-showcase-datetime-fixture @ ab36336e7 (pushed 2026-08-08 by the round-2 dev, whose session had write here).
• Its two red checks are both diagnosed, with the fix ready:
  1. check:i18n-coverage (inside the "TypeScript Type Check" job): the fixture's label: 'Incurred At' grew the example's frozen i18n debt 451 → 452. Fix = exactly ONE translated field (en Incurred At / zh-CN 发生时间) — deliberately one, because translating more drops the count below the 451 baseline and the same gate rejects the un-ratcheted improvement. The verbatim diff is posted as a comment on PR test(showcase): 给内嵌网格子对象补 datetime 永久夹具 (objectui#3569 配套) #7328 (comment id 5237329627); reverse-verified (removing it reproduces the CI error verbatim; applying it, check-i18n-coverage: OK).
  2. Check Changeset: resolved already — skip-changeset label applied (the gate's own preferred route for examples/-only PRs); the relabel run shows the check skipped. It clears fully on the next push.
• Also verified with the diff applied on a merge with current main (clean, zero conflicts): full workspace build 66/66, showcase validate ✓, typecheck clean, 15 files / 158 tests green, check:i18n OK, control-byte scans clean.

What this lane needs to do (mechanical)

git fetch origin main claude/issue-3569-showcase-datetime-fixture
git checkout claude/issue-3569-showcase-datetime-fixture && git merge origin/main   # clean
# apply the diff from PR #7328 comment 5237329627
git commit -am "test(showcase): translate incurred_at label" 
git push origin claude/issue-3569-showcase-datetime-fixture

Then merge #7328 once green. If anything surprises, ping the objectui seat thread on objectui#3569 rather than re-deriving — the full provenance is there.

Known-and-filed neighbor, not part of this: #6533 (time control has unit tests but no live fixture).

[claude]

Triage: routed domain:engine-core (kept pm:queue).

• What it is: mechanical adoption of PR test(showcase): 给内嵌网格子对象补 datetime 永久夹具 (objectui#3569 配套) #7328 — one prepared i18n commit (diff posted as PR comment 5237329627, reverse-verified by the filing seat) + merge once green. Verified at triage time: PR test(showcase): 给内嵌网格子对象补 datetime 永久夹具 (objectui#3569 配套) #7328 still open, head ab36336e7 unchanged, branch alive.
• Routing rationale: the landing files are examples/app-showcase (the permanent showcase_expense_line.incurred_at datetime fixture) plus one i18n pair. Per "examples/** belongs to the subsystem it exercises": the fixture pins the datetime value round-trip on the platform data path — the repro data behind objectui#3569's embedded sub-table datetime loss — ⇒ domain:engine-core (same family as A CEL defaultValue stores a raw Date into datetime/date fields — the stored-value contract says ISO-8601 string, and the NOW() token normalizes but the CEL branch does not #7373's datetime-value handling). If the lane reads the landing differently, flag per the mislabel path rather than relabeling.
• Dedup: this card IS the single dispatch entry for PR test(showcase): 给内嵌网格子对象补 datetime 永久夹具 (objectui#3569 配套) #7328; objectui#3569 closed via objectui#4083. [showcase] 内嵌子表 time 列没有真机夹具：两个 inlineEdit 网格都无自然落脚点，且都已顶到 6 列预算 #6533 (time control live fixture) is a named neighbor, not a dup. Clean.

本评论来自分诊座位 Routine（#5474 试点），不构成认领。

---

Generated by Claude Code

[claude]

Claim: PM loop wave 8 (seat #6019) — seat-executed, S-grade mechanical adoption
Session: session_01MwoubC3jL271FYt9rGXwxb
Branch: claude/issue-3569-showcase-datetime-fixture (adopting PR #7328's existing branch, per the handoff recipe)
Worktree: objectstack-7358
Domain: domain:engine-core

Race re-read: handoff body + triage comment re-read; no competing session. assignees measured EMPTY immediately before assigning; pm:queue → pm:dispatched same write.

Executing the card's own recipe: merge origin/main into the branch, apply the reverse-verified i18n diff from PR #7328 comment 5237329627, push, merge on green. ⛔ Nothing re-derived — provenance stays with objectui#3569 as instructed.

---

Generated by Claude Code

[claude]

Executed — pushed 94161354b, auto-merge armed on PR #7328.

Followed the recipe with one deviation and one environment note, both worth recording:

1. The diff block in comment 5237329627 did not git apply (markdown-mangled hunk); the entries were applied by anchored content replacement instead, byte-faithful to the prepared text (23 insertions: both comment blocks + showcase_expense_line.incurred_at in en and zh-CN). Verified against the comment's own text, not re-derived.
2. This clone is shallow — git merge origin/main first failed with refusing to merge unrelated histories; --deepen=200 found the ancestor and the merge is clean (picks up today's scripts/pm/ tooling among others).
3. Local check:i18n-coverage requires a built workspace CLI (its own PREREQUISITE NOT MET message is admirably loud), so the local gate was not re-run — the entry was already reverse-verified in both directions by the filing seat on a full build, and CI on this push is the arbiter. skip-changeset label already in place clears the changeset red on this run.

If CI comes back green, #7328 merges on its own and this card + objectui#3569's last piece close out. If the i18n gate is still red on this push, that is a new fact the handoff didn't predict — I will report on the objectui#3569 thread per the card's instruction rather than re-deriving.

---

Generated by Claude Code

[claude]

Done — PR #7328 merged as ec0f871fd; closing completed.

The adoption ran exactly as the handoff predicted once the prepared i18n entry was on the branch: skip-changeset cleared the changeset gate on the new run, check:i18n-coverage came back green in CI (the entry's both-directions verification held), and auto-merge carried it through the queue. objectui#3569's last remaining piece is landed; the permanent showcase datetime fixture (showcase_expense_line.incurred_at) is on main.

Closing manually — the PR predates this card, so its body carries no Closes #7358.

---

Generated by Claude Code