agentleFS
207studio207studio/jev-claude-tools/skills/jev-code-sweep/SKILL.md

jev-code-sweep

Skill1 starsChanged 53 days agoMITView on GitHub

Sweep a whole repository or large change for likely runtime defects by having Jev triage every ~100-line chunk before any agent reads code, then send only the top clusters to reviewers. Use for "review the whole codebase", pre-release defect sweeps, or when a review would otherwise load many files into context.

The file

Files
SKILL.md

Jev 우선 코드 검수 스윕

저장소 전체를 에이전트가 읽기 전에 Jev가 청크마다 결함 가능성을 판정하고, 상위만 검수 에이전트에게 넘긴다. 파일 몇 개·알려진 버그 하나를 볼 때는 쓰지 않는다 — 그건 그냥 읽는다.

도구는 ~/.claude/jev/에 있다: chunk.py(청킹), jevq.py(코드 슬라이스 판정·주장 대조). 둘 다 코드 본문을 출력하지 않는다.

  1. 청킹 → 2. jev-mode batch로 1차 판정 → 3. 코드가 확률 조합·임계값 → 4. 상위를 묶어 검수 에이전트에 배정 → 5. 발견마다 jevq claim + 검증자 반박, 결합은 코드가. 명령·임계값·질문은 레시피, 질문은 1차·2차.

검수 에이전트에게 Jev 사용을 산문으로 지시하지 않는다. 출력 스키마의 required 필드로 강제한다 — jev-mode의 위임 규칙. 에이전트 모델은 ~/.claude/jev/agent-tier.json으로 정한다. 부모가 Fable/Opus면 상속시키지 않는다.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.