daily-fix: follow-up defer/teardown must re-park parent
Highlight text on any card (body, plan, or an event) to anchor a comment, or leave a whole-task comment here. Mention @claude to summon a reply.
No comments yet.
Overview / Motivation
Auto-filed by the /daily 2026-07-14 problem sweep (session 09f28ede, #825): the wedged naturalistic Track-S run 2 was torn down and the round deferred, but the parent stayed at followups_running ~3.5h (deferred ~00:30Z, re-parked by the 01:52Z tick).
Goal
add one line to the same-issue follow-up loop: any mid-round defer/teardown immediately restores the parent to its pre-round resting state (awaiting_promotion) in the same action as the pod/instance teardown
Workflow gap
- Bug observed: a mid-round defer/teardown of a same-issue follow-up left #825 stuck at followups_running for ~3.5h with no live compute — the teardown step has no re-park duty, only the 45-min issue-tick caught it
- Why it is a workflow gap: the follow-up loop's SKILL.md text holds the task at
followups_runningfor the round but names no status-restore duty on the defer/teardown path; the tick/watcher are backstops, not the owner. - Confidence: high (incident observed; the backstop, not the loop, restored status)
- verified-at-filing:
grep -n "defer" .claude/skills/issue/SKILL.md-> hits only in unrelated contexts (concern_deferral_request, deferred tools); no mid-round defer/teardown re-park line exists (absence-of-guard claim; 0 relevant in-target hits) (2026-07-15).
Proposed change
One line in SKILL.md § Same-issue follow-up loop: "A mid-round defer / teardown (wedge, pathological fit, user defer) restores the parent to its pre-round resting status in the SAME action as the teardown — never leave followups_running without live round compute."
Constraints
- Workflow-surface only;
workflow_lint.py --check-askspasses; recursion guard applies.
Provenance
- workflow_fix_target: .claude/skills/issue/SKILL.md
- fingerprint: b3abe37a4394