Filed unassigned by the dev seat that codified the strip-on-close convention (PR #12906). Small, but it is a protocol document disagreeing with another protocol document about a rule a patrol now enforces.
The two sentences
PR #12906 adds one sentence to .claude/skills/pm-dispatch/SKILL.md's label-discipline section:
关闭即摘 pm:* 状态标(与关单同一笔;domain:*/类型标签留下 —— 归属不是状态)。
references/landing-operations.md section B still says, of the same moment:
确认 MERGED 的同一动作里给 Part of 卡收口(Fixes 卡 GitHub 代关、标签随卡离开在飞视图)
The parenthetical is the older reading: a Fixes card is auto-closed by GitHub, and its pm:* label is treated as leaving the in-flight view with the card — true only because the inventory queries filter is:open. Under the codified convention the label does not merely stop being counted, it is stripped, and the closed-card patrol row (H22) reports it when it is not.
So the landing paragraph now describes the Part of case correctly (strip the label, comment, one action) while implying the Fixes case needs no label write at all — which is exactly the half the convention just closed.
Why it is not a rider on #12906
It is a second file, outside that PR's declared file surface, and the fix is a judgement rather than a mechanical respelling: the parenthetical is explaining why the in-flight view stays honest, and that explanation is still partly true. Whoever takes it has to decide whether to correct it, delete it as now-redundant, or keep it and add the strip step — and landing-operations.md is governed and ratcheted, so a net-0 edit with a cut ledger is needed either way.
Scope note
⛔ Not a request to backfill labels on already-closed cards. The patrol's new dated closure floor means historical closures are out of scope by construction; this is only about what the document tells the next seat to do at the next merge.
Refs: PR #12906 (the convention sentence) · objectstack-ai/objectui#5985 (the ruling it executes)
Generated by Claude Code
Filed unassigned by the dev seat that codified the strip-on-close convention (PR #12906). Small, but it is a protocol document disagreeing with another protocol document about a rule a patrol now enforces.
The two sentences
PR #12906 adds one sentence to
.claude/skills/pm-dispatch/SKILL.md's label-discipline section:references/landing-operations.mdsection B still says, of the same moment:The parenthetical is the older reading: a
Fixescard is auto-closed by GitHub, and itspm:*label is treated as leaving the in-flight view with the card — true only because the inventory queries filteris:open. Under the codified convention the label does not merely stop being counted, it is stripped, and the closed-card patrol row (H22) reports it when it is not.So the landing paragraph now describes the
Part ofcase correctly (strip the label, comment, one action) while implying theFixescase needs no label write at all — which is exactly the half the convention just closed.Why it is not a rider on #12906
It is a second file, outside that PR's declared file surface, and the fix is a judgement rather than a mechanical respelling: the parenthetical is explaining why the in-flight view stays honest, and that explanation is still partly true. Whoever takes it has to decide whether to correct it, delete it as now-redundant, or keep it and add the strip step — and
landing-operations.mdis governed and ratcheted, so a net-0 edit with a cut ledger is needed either way.Scope note
⛔ Not a request to backfill labels on already-closed cards. The patrol's new dated closure floor means historical closures are out of scope by construction; this is only about what the document tells the next seat to do at the next merge.
Refs: PR #12906 (the convention sentence) · objectstack-ai/objectui#5985 (the ruling it executes)
Generated by Claude Code