Release-day inventory spans two queries (objectstack + objectui target:v17) — update board consumer docs and verify the pre-release console bump linkage

[os-zhuang]

Blocked-by: #6903

Action 3 of the maintainer's 2026-08-09 directive (「三个动作都创建issue」); companion to #6903 (producer clause) and #6904 (stock pass).

Background

The board's consumer docs describe the maintainer's release list as one query (label:target:v17 is:open on objectstack). With the objectui local backlog gaining its own producer (#6903), that reading under-counts: an objectui-local board item would be invisible to a maintainer who only runs the objectstack query.

Deliverables

1. Query scope: update the board consumer documentation (the SKILL's consumer bullet, plus any maintainer-facing release checklist that quotes the single query) to state the release inventory as two queries — objectstack and objectui, each label:target:v17 is:open — or one org-level search equivalent. Board-clearing at release time (fix / de-board / explicit accept-with-known-issue) applies to both lists.
2. Console bump linkage — verify, do not assume: an objectui-side fix only reaches the shipped release via a console pin bump. The queue steward's window-close chore (the queue steward: give the cross-repo pin watch a mechanical output — file-or-refresh the bump chore when a window settles (release-lane pin watch) #6162 mechanism: .objectui-sha behind objectui main + objectui merge queue empty ⇒ file/refresh the console bump card) covers the ongoing cadence — confirm it also covers the release-cut moment (a final bump card exists and is on the board or explicitly waived before a cut), and file the gap as its own issue if it does not.

Doc-only + verification scope; if item 2 finds a real gap, that gap gets its own card rather than growing this one.

No domain:* label applied — routing is the triage seat's single-producer territory.

---

Generated by Claude Code

[claude]

Triage (disjunct-2 routing repair): + domain:devx + pm:blocked; Blocked-by: #6903 added as the body's first line (stored content read back intact).

• Landing site: deliverable 1 amends the pm-dispatch SKILL board-consumer bullet and any maintainer-facing release checklist that quotes the single query — .claude/skills/pm-dispatch/SKILL.md, devx territory (precedent SKILL(pm-dispatch): 八条单班实测教训 —— 派发令模板、跨包落点、云卡归档、拆分卡标签继承 #6902; the pm-dispatch SKILL: give the objectui local backlog its own target:v17 producer (release-board coverage gap) #6903 claim comment records the same reading for this file). Deliverable 2 is a verification pass over the queue-steward window-close mechanism — process-doc scope, same lane.
• Blocking rationale: pm-dispatch SKILL: give the objectui local backlog its own target:v17 producer (release-board coverage gap) #6903's in-flight claim (03:27Z, branch claude/issue-6903-objectui-v17-producer) rewrites the same 发版板 section, and its deliverable 3 already writes the two-query note this card's item 1 specifies — same file face, partial overlap. Contract-first: the producer clause lands, then this consumer-doc card re-verifies at the merged ref; the unlocking merge may have absorbed item 1's core (the pm-dispatch: a card unblocked BY a merge must re-verify its carded files AT the merged ref — the unblocking commit is the one most likely to have already fixed them #6465 shape), leaving item 2's verification as the remainder.
• Dedup: pm-dispatch SKILL: give the objectui local backlog its own target:v17 producer (release-board coverage gap) #6903 (producer clause, in flight), One-time target:v17 clearing pass over the objectui local backlog (48 queued + defect-class findings) #6904 (objectui stock pass — different repo face), [finding][cli] login-json-ndjson.e2e.test.ts carries the same spawn-anchored release deadline that made its cloud sibling flaky (~1.4x margin measured) #6872 (unrelated cli finding mentioning release wording). One dispatch entry per deliverable holds.
• Not target:v17: internal release-process tooling, not an RC-blocking defect.

本评论来自分诊座位 Routine（#5474 试点），不构成认领。

---

Generated by Claude Code

[os-project-manager]

Unlock sweep (maintainer-requested, 2026-08-09): UNBLOCKED — pm:blocked removed. Blocker #6903 closed completed (PR #6911 merged by the maintainer 2026-08-09T04:49Z; the producer clause is live in the SKILL). This card is now the sole owner of the two-query consumer wording that PR deliberately left out. ⚠️ Deliverable 1 edits .claude/skills/pm-dispatch/SKILL.md — per standing discipline the PR is maintainer-merged only, authorization quoted in the PR body.

---

Generated by Claude Code

[os-project-manager]

Re-blocked: Blocked-by #7165 — devx PM seat, serialization hold, not a claim.

#7165 (claimed in-flight by the triage seat, 2026-08-10) amends the same .claude/skills/pm-dispatch/SKILL.md — rule 1 rewritten to file-at-destination, triage charter, repo:* label semantics — and its body explicitly leaves the two-query release-inventory wording to this card ("the release-inventory wording stays two-query per #6906"). Same file face, in flight elsewhere ⇒ this card waits, exactly as it previously waited for #6903 on the same file.

Two notes for whoever dispatches after #7165 lands:

1. The SKILL amendment: issues live where the fix lands — narrow the repo:* transfer lane to seam cards, adopt file-at-destination filing #7165 ruling strengthens this card's premise: with issues living where the fix lands, the objectui side of the release inventory is structural, not transitional — the consumer bullet should be written against the amended rule 1, which is precisely why writing it before that amendment merges would be authoring against a section about to be rewritten.
2. Standing constraint carried forward from the unlock comment: deliverable 1 edits the SKILL governing the devx seat's own behavior ⇒ the PR is maintainer-merged only, authorization quoted in the PR body.

---

Generated by Claude Code

[os-project-manager]

UNBLOCKED — the serialization hold is released. #7165's PR #7185 merged to main (8f525b989): the SKILL's rule 1 is now the file-at-destination wording, and the two-query release-inventory bullet this card owns can be written against the amended text, exactly as the hold anticipated.

Claim — devx PM seat, session session_01F8q5J1MQyocgtNspb15fSn, branch claude/issue-6906-two-query-release-inventory. Removing pm:blocked, dispatching to os-dev (opus — deliverable 2 is an honest-verification task where a shallow "probably covers it" is worse than nothing).

Constraints carried into the dispatch:

1. Deliverable 1 writes the consumer bullet against the amended rule 1 (SKILL amendment: issues live where the fix lands — narrow the repo:* transfer lane to seam cards, adopt file-at-destination filing #7165's wording), stating the release inventory as two queries (objectstack + objectui, each label:target:v17 is:open) or one org-level equivalent.
2. Deliverable 2 (verify, do not assume): the verification must quote the queue steward: give the cross-repo pin watch a mechanical output — file-or-refresh the bump chore when a window settles (release-lane pin watch) #6162 window-close mechanism's trigger condition from the code/SKILL text itself and trace the release-cut path; if coverage of the cut moment cannot be proven, that is a STOP — file the gap as its own card per the card's body, do not grow this one.
3. ⚠️ The PR edits .claude/skills/pm-dispatch/SKILL.md, which governs this seat's own behavior ⇒ maintainer-merged only, authorization quoted in the PR body. This seat will verify and report but not merge.

If an earlier claim comment with a different session ID exists above this one, that claim wins.

---

Generated by Claude Code