workflow-fix: follow-up brief marker versions defer to max+1 rule

kind: infra#wf-fix#wf-fix-fp:bb0e9be7e388

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 workflow-fix-on-bug protocol from a workflow-fix candidate raised on task #825 (emitting agent: experiment-implementer).

Goal

The same-issue follow-up loop's implementer brief guidance should defer marker versions to the standing max+1 posting rule instead of hardcoding per-round version numbers.

Workflow gap

Bug observed: The same-issue follow-up loop's implementer brief instructed "post as epm:experiment-implementation v1" on a task whose marker sequence was already at v6, causing a version collision the implementer had to detect and re-post around.
Why it is a workflow gap: The brief template hardcodes per-round version numbers instead of deferring to the standing max+1 posting rule, so every follow-up round on a task with prior implementation markers reproduces the #389 collision class.
Confidence (emitter): high

Proposed change (candidate diff sketch — refine in planning)

- post as `epm:experiment-implementation v1` via ...
+ post as `epm:experiment-implementation v<max_existing+1>` (read events.jsonl for the
+ highest existing version of the marker key first; task.py post-marker does NOT
+ auto-increment) via ...

Scope / surfaces

Primary target: .claude/skills/issue/SKILL.md
Grep the workflow surface for hardcoded per-round marker-version instructions in brief templates (grep -rn 'epm:experiment-implementation v1' .claude/ CLAUDE.md) and update every hit; list them in the plan. Consider whether the Step 4b brief template and the same-issue follow-up loop § step 3 both need the max+1 wording.

Constraints / invariants

Workflow-surface only — never experiment code, configs/, or tasks/.
scripts/workflow_lint.py --check-asks passes; ruff on touched files passes.
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/skills/issue/SKILL.md
fingerprint: bb0e9be7e388

target_file: .claude/skills/issue/SKILL.md bug_observed: The same-issue follow-up loop's implementer brief instructed "post as epm:experiment-implementation v1" on a task whose marker sequence was already at v6, causing a version collision the implementer had to detect and re-post around. why_workflow_gap: The brief template hardcodes per-round version numbers instead of deferring to the standing max+1 posting rule, so every follow-up round on a task with prior implementation markers reproduces the #389 collision class. proposed_change: The follow-up-loop brief should say "post at the next version after the highest existing epm:experiment-implementation row in events.jsonl" rather than a literal version. diff_sketch: |

post as epm:experiment-implementation v1 via ...

post as epm:experiment-implementation v<max_existing+1> (read events.jsonl for the
highest existing version of the marker key first; task.py post-marker does NOT
auto-increment) via ... confidence: high related_task: #825

Activity