Structured review findings with auto gap registration#4
Merged
Conversation
- Tiered extraction: <findings> tag, bare JSON array, code block, graceful empty - _repair_json helper: strips fences, trailing commas, single quotes (stdlib-only) - SEVERITY_TO_PRIORITY mapping: critical→required, major→important, minor/nitpick→nice-to-have - --register flag auto-adds critical/major findings as gaps via cmd_gap_add - Unknown severity defaults to "important" with warning - Cap at 50 findings max - 5 smoke tests: valid tag, missing tags, malformed JSON, --register, gap verification Task: fn-3-structured-review-findings-auto-gap-add.1
- Add optional <findings> schema to plan-review, impl-review, and epic-review RP workflow.md Output Format sections - Add same schema to build_review_prompt (impl + plan) and build_completion_review_prompt (epic) Codex prompts in prompts.py - Schema is consistent across all 6 locations: title, severity, location, recommendation fields - Findings are optional — SHIP reviews with no issues may omit Task: fn-3-structured-review-findings-auto-gap-add.2
- Plan-review Codex + RP fix loops call parse-findings --register --source plan-review - Impl-review Codex + RP fix loops call parse-findings --register --source impl-review - Epic-review Codex + RP fix loops call parse-findings --register --source epic-review - All 6 fix loops log the number of findings registered as gaps - Add parse-findings to docs/flowctl.md Available Commands list - Add ### parse-findings section with usage, options, extraction strategy, severity mapping Task: fn-3-structured-review-findings-auto-gap-add.3
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Closes the loop between review findings and the gap registry. Previously review skills output issues as prose — now they produce structured
<findings>JSON that's automatically parsed and registered as tracked gaps.flowctl parse-findings: Extracts<findings>JSON from review output with tiered fallback (tag → regex → none). Optional--registerflag auto-callsgap addwith severity→priority mapping.<findings>schema in all 6 review prompts: Added to RP workflow.md and Codex prompts.py for plan-review, impl-review, and epic-review.parse-findings --registerwired into all 6 fix loop paths (Codex + RP × 3 review types).parse-findingsdocs indocs/flowctl.md.Test plan
python3 -m pytest scripts/flowctl/tests/— 255 passedbash scripts/smoke_test.sh— 95/95 passed (5 new parse-findings tests)🤖 Generated with Claude Code