feat(me/evolve): clean indexer, add summary view, English SKILL.md - #665
Conversation
Signal: user asked 'apply-patch.sh는 왜쓰는거야?' at turn 867 Session: 7f6cbce3-7166-4ccb-8c1e-0a70f17e7ecb
This reverts commit 7b9e49f.
Phase 2에서 메인 에이전트가 git apply + git commit을 직접 수행한다. 외부 캐시 차단은 SKILL.md 지시문만으로 강제.
- Filter structural noise (Base directory, <bash-*>, [Request interrupted...]) — user count 64 → 41 on sample
- Exclude bookkeeping tools (Task*, AskUserQuestion) from prior so classifier sees real assistant actions
- Add summary {headline, clusters, signal_positions} for shallow-then-deep exploration
- SKILL.md: read summary first, drop stale `groups` reference, spec "nearest AGENTS.md/CLAUDE.md" resolution
Output now has only session_id, session_title, turns, summary, events. Removed tools_top, skill_runs, signal_counts, --skill filter, filterBySkill. summary.headline still carries the same counts the LLM needs.
description triggers keep Korean phrases (스킬 개선 / 회고) since they are user-facing invocation keywords.
|
Warning Review limit reached
More reviews will be available in 55 minutes and 4 seconds. Learn how PR review limits work. Your organization has run out of usage credits. Purchase more in the billing tab. ⌛ How to resolve this issue?After more reviews become available, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available. Please see our Fair Usage Limits Policy for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughThis PR refactors the ChangesEvolve Skill Refactoring
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~30 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Replace tools_top / skill_runs / --skill / signal_counts tests with top-level key guard and summary structure assertions.
# [17.16.0](v17.15.0...v17.16.0) (2026-05-27) ### Features * **me/evolve:** clean indexer, add summary view, English SKILL.md ([#665](#665)) ([b89c5a6](b89c5a6))
Summary
summary+events(droppedtools_top,skill_runs,signal_counts, and the--skillfilter)summaryfield withheadline/clusters/signal_positionsfor shallow-then-deep explorationpriorso the classifier sees real assistant actionsapply-patch.shhelper; main agent now runsgit apply+git commitdirectly per SKILL.mdTest plan
bun build-index.ts <sample.jsonl>→ exit 0, output keys = {events, session_id, session_title, summary, turns}signal_counts.user64 → 41 (pseudo-user filtered)priorno longer contains TaskUpdate/TaskCreate/etc as primary actionsummary.headline: "868 turns · 41 users · 10 interrupts · 2 errors · 7 repeats · 5 clusters"summary.clusterssurfaces 5 friction regions on sampleSummary by CodeRabbit
Documentation
/me:evolveskill instructions with clearer control-flow definitions, invocation modes (including--dry-run), and constraints around transcript and project handling.Refactor
Tests