EPS
← All tasks·#1188Completed

workflow-fix: analyzer re-fold H1 retitle set-title sync

kind: infra#wf-fix#wf-fix-fp:a38bd2240cf0#daily-auto-filed
track:

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 Step-C parked-candidate sweep (2026-07-08) from a candidate parked on task #1110 (recursion-guarded workflow-fix session).

Goal

Add 'then run task.py set-title <N> "<new H1 text>"' to the analyzer's same-issue follow-up re-entry step (mirror the main promotion sequence's set-body-then-set-title ordering); also update the sibling sites the candidate greps named in .claude/skills/issue/SKILL.md (~:6033-6037, ~:6234-6238).

Workflow gap

  • Bug observed: analyzer.md's same-issue follow-up re-entry step licenses retitling the clean-result H1 followed by set-body only — set_body preserves the frontmatter title, so every headline-moving follow-up round produces an H1/frontmatter divergence the #1110 verifier check FAILs at the next gate (a true positive, but a guaranteed one-bounce round-trip).
  • Why it is a workflow gap: analyzer.md prescribes the H1 edit without the paired task.py set-title sync, so the two title surfaces diverge by construction on every headline-moving re-fold.
  • Confidence (emitter): high

Proposed change (candidate diff sketch — refine in planning)

In analyzer.md § Same-issue follow-up re-entry:

  • ... retitle the H1 if the headline moved, then task.py set-body <N> --file ...
  • ... retitle the H1 if the headline moved, then task.py set-body <N> --file ...,
  • then `task.py set-title <N> "<new H1 text>"` (keeps frontmatter == H1;

  • the #1110 verifier check FAILs the gate otherwise)

Scope / surfaces

  • Primary target: .claude/agents/analyzer.md
  • Grep the workflow surface for the pattern before editing (grep -rln '<pattern>' .claude/ CLAUDE.md scripts/) and update every hit; list them in the plan.
  • Sibling sites: .claude/skills/issue/SKILL.md follow-up paths (candidate named ~:6033-6037 and ~:6234-6238; re-grep for set-body near the follow-up fold steps).

Constraints / invariants

  • Workflow-surface only — never experiment code, configs/, or tasks/.
  • scripts/workflow_lint.py --check-asks passes; ruff on touched files passes; if workflow.yaml or CLAUDE.md change, they stay consistent with the rule file.
  • This session runs under EPM_WORKFLOW_FIX_SESSION=1 and carries a workflow_fix_target: Provenance line — it MUST NOT auto-route any of its own subagents' workflow-fix candidates (recursion guard).

Provenance

  • workflow_fix_target: .claude/agents/analyzer.md
  • origin: parked candidate on task #1110 at 2026-07-07T13:04:28Z

Verbatim parked note:

parked — running under workflow_fix_target Provenance (recursion guard, .claude/rules/workflow-fix-on-bug.md § Recursion guard); LOGGED for the next orchestrator/human pass, NOT auto-filed.

target_file: .claude/agents/analyzer.md bug_observed: The same-issue follow-up re-entry step licenses the analyzer to retitle the clean-result H1 ("retitle the H1 if the headline moved", analyzer.md ~:446-456) followed by set-body only — set_body preserves the frontmatter title, so every headline-moving follow-up round now produces an H1/frontmatter divergence that the new #1110 verifier check FAILs at the next gate (a true positive, but a guaranteed one-bounce round-trip). why_workflow_gap: analyzer.md prescribes the H1 edit without the paired task.py set-title sync, so the two title surfaces diverge by construction on every headline-moving re-fold. proposed_change: Add "then run task.py set-title """ to the analyzer's same-issue follow-up re-entry step (and its main promotion sequence already does set-body then set-title — mirror that ordering). diff_sketch: | In analyzer.md § Same-issue follow-up re-entry:

  • ... retitle the H1 if the headline moved, then task.py set-body <N> --file ...
  • ... retitle the H1 if the headline moved, then task.py set-body <N> --file ...,
  • then `task.py set-title <N> "<new H1 text>"` (keeps frontmatter == H1;

  • the #1110 verifier check FAILs the gate otherwise)

confidence: high related_task: #1110

(Origin: codex-critic alternatives-lens Must-Fix on the #1110 plan, demoted to a non-blocking recommendation by the binding reconciler verdict — facts verified: set_body preserves fm title; analyzer.md + issue SKILL.md follow-up paths name only set-body. Also grep candidate sites: .claude/skills/issue/SKILL.md:6033-6037, :6234-6238.)

Activity