Skip to content

feat(me/evolve): clean indexer, add summary view, English SKILL.md - #665

Merged
baleen37 merged 7 commits into
mainfrom
fancy-quartz-turing
May 27, 2026
Merged

feat(me/evolve): clean indexer, add summary view, English SKILL.md#665
baleen37 merged 7 commits into
mainfrom
fancy-quartz-turing

Conversation

@baleen37

@baleen37 baleen37 commented May 27, 2026

Copy link
Copy Markdown
Owner

Summary

  • Indexer now outputs only summary + events (dropped tools_top, skill_runs, signal_counts, and the --skill filter)
  • Added summary field with headline / clusters / signal_positions for shallow-then-deep exploration
  • Cleaned pseudo-user noise (Base directory tags, bash IO blocks, [Request interrupted...]) → user-event count 64 → 41 on sample transcript
  • Excluded bookkeeping tools (Task*, AskUserQuestion) from prior so the classifier sees real assistant actions
  • Removed apply-patch.sh helper; main agent now runs git apply + git commit directly per SKILL.md
  • Translated SKILL.md body to English (kept Korean trigger phrases in description)

Test plan

  • bun build-index.ts <sample.jsonl> → exit 0, output keys = {events, session_id, session_title, summary, turns}
  • signal_counts.user 64 → 41 (pseudo-user filtered)
  • prior no longer contains TaskUpdate/TaskCreate/etc as primary action
  • summary.headline: "868 turns · 41 users · 10 interrupts · 2 errors · 7 repeats · 5 clusters"
  • summary.clusters surfaces 5 friction regions on sample
  • SKILL.md body has 0 Korean characters; frontmatter description triggers preserved
  • 12/12 BATS tests pass

Summary by CodeRabbit

  • Documentation

    • Refined /me:evolve skill instructions with clearer control-flow definitions, invocation modes (including --dry-run), and constraints around transcript and project handling.
  • Refactor

    • Updated session indexing to produce simplified output with summary information (headlines, clusters, signal positions).
    • Improved patch application workflow with direct git commands and per-change commits; removed intermediate helper script.
    • Strengthened external cache validation and safety guardrails.
  • Tests

    • Updated test suite to reflect changes to indexing and patch workflow.

Review Change Stack

Test User added 6 commits May 27, 2026 22:17
Signal: user asked 'apply-patch.sh는 왜쓰는거야?' at turn 867
Session: 7f6cbce3-7166-4ccb-8c1e-0a70f17e7ecb
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.
@baleen37
baleen37 enabled auto-merge (squash) May 27, 2026 14:23
@coderabbitai

coderabbitai Bot commented May 27, 2026

Copy link
Copy Markdown

Warning

Review limit reached

@baleen37, we couldn't start this review because you've reached your PR review rate limit.

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 @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

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 configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 80a89e10-f7da-4825-9de9-4ee1a925ee60

📥 Commits

Reviewing files that changed from the base of the PR and between 23dd6d2 and 20bb55c.

📒 Files selected for processing (1)
  • tests/me/evolve-build-index.bats
📝 Walkthrough

Walkthrough

This PR refactors the /me:evolve skill by rewriting its documentation, restructuring the session indexing pipeline in build-index.ts, removing the apply-patch.sh helper script, and updating corresponding tests. The workflow transitions from apply-patch.sh-mediated patching to direct git apply + git commit operations, with explicit Phase 0–2 specifications and stricter external-cache handling.

Changes

Evolve Skill Refactoring

Layer / File(s) Summary
SKILL.md workflow specification
plugins/me/skills/evolve/SKILL.md
SKILL.md is rewritten in English with detailed invocation rules and three-phase workflow: Phase 0 calls build-index.ts [--session <id>] to build the session index; Phase 1 specifies LLM-only event classification and structured proposal generation routed to SKILL/AGENTS/CLAUDE; Phase 2 replaces the apply-patch.sh helper with direct git apply + git commit per patch and routes external-cache proposals to dated upstream-suggestions.md without editing.
build-index.ts data model and indexing pipeline
plugins/me/skills/evolve/scripts/build-index.ts
build-index.ts introduces Cluster and Summary interfaces; the SessionIndex now includes a summary object with headline, clusters, and signal positions, while prior fields tools_top, skill_runs, signal_counts are removed. Adds BOOKKEEPING_TOOLS filter to exclude bookkeeping tool uses from prior assistant actions; introduces PSEUDO_USER_PREFIXES and isPseudoUser() to suppress pseudo-user inputs from user events. The indexing pipeline is refactored via buildClusters, buildSignalPositions, buildHeadline, buildSummary builders. The --skill CLI option and skill-filter parameter are removed from buildIndex and all call sites.
Test suite updates for apply-patch.sh removal
tests/me/evolve-skill.bats, tests/me/me-specific.bats
Assertions for apply-patch.sh file existence and executability are removed from test suites; the dedicated test verifying apply-patch.sh blocks external cache writes is deleted. The me: evolve skill exists test now validates scripts/build-index.ts instead of scripts/apply-patch.sh.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~30 minutes

Possibly related PRs

  • baleen37/bstack#664: Refactors /me:evolve implementation by changing build-index.ts output schema and removing the apply-patch.sh helper script in favor of direct git apply+git commit workflow.

Poem

🐰 Hop and refactor, the indexing flows anew,
Bookkeeping filtered, pseudo-users too,
Apply-patch fades; git commits take the stage,
Phase 0, 1, 2—a clearer evolving page!

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Docstring Coverage ⚠️ Warning Docstring coverage is 16.67% which is insufficient. The required threshold is 80.00%. Write docstrings for the functions missing them to satisfy the coverage threshold.
✅ Passed checks (4 passed)
Check name Status Explanation
Description Check ✅ Passed Check skipped - CodeRabbit’s high-level summary is enabled.
Title check ✅ Passed The title clearly and concisely summarizes the three main changes: indexer cleanup (removing unused fields), new summary view feature, and English translation of SKILL.md documentation.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.

✏️ Tip: You can configure your own custom pre-merge checks in the settings.

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch fancy-quartz-turing

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.

❤️ Share

Comment @coderabbitai help to get the list of available commands and usage tips.

Replace tools_top / skill_runs / --skill / signal_counts tests with
top-level key guard and summary structure assertions.
@baleen37
baleen37 merged commit b89c5a6 into main May 27, 2026
4 of 5 checks passed
@baleen37
baleen37 deleted the fancy-quartz-turing branch May 27, 2026 14:25
baleen-release-app Bot pushed a commit that referenced this pull request May 27, 2026
# [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))
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant