Pre-action boundary checking — validates agent tool calls against declared capabilities and task contracts
日本語の概要は準備中です。原文の説明を表示しています。
Use when encountering any bug, test failure, or unexpected behavior. Enforces a strict reproduce-first, root-cause-first, failing-test-first debugging workflow before fixing.
インストール方法を見るインストールする前に、エージェントに与えられる指示の中身を確認できます。
엄격한 디버깅 워크플로우다. 버그, 테스트 실패, 예기치 않은 동작을 다룰 때 사용한다.
핵심 목적은 세 가지다.
다음 규칙은 예외 없이 따른다.
이 과정을 어기는 것은 디버깅 실패로 본다.
다음 상황이면 이 스킬을 사용한다.
다음 핑계는 허용하지 않는다.
이 스킬을 사용할 때는 내부적으로 아래 항목을 반드시 고정한다.
이 7개 중 빠진 항목이 있으면 아직 끝난 일이 아니다.
반드시 아래 순서로 진행한다.
Trigger: When any external dependency, tool, or resource appears unavailable.
Before declaring a task blocked or unsolvable, exhaust this checklist:
| # | Workaround Path | Check |
|---|---|---|
| 1 | Environment bypass | Can the dependency be bypassed in the current environment? |
| 2 | Alternative tool | Is there an alternative tool, library, or approach? |
| 3 | Partial solution | Can the task be partially completed without the dependency? |
| 4 | Install/configure | Can the dependency be installed or configured now? |
| 5 | Existing credentials | Are related tokens, auth files, or configs already present? |
Rules:
Output format:
[Blocker Triage]
├── Dependency: {what is unavailable}
├── Path 1: {attempted} → {outcome}
├── Path 2: {attempted} → {outcome}
├── Path 3: {attempted} → {outcome}
└── Verdict: {proceed with workaround N | genuinely blocked — reason}
If verdict is "genuinely blocked", escalate to user with full triage report. Do NOT silently abandon the task.
먼저 문제를 축약한다.
출력 형식:
Problem: <expected> but got <actual> under <condition>
증상과 추측을 섞지 않는다.
Good: Product detail API returns 500 when brand is null.
Bad: Serializer is broken because brand mapping seems wrong.
수정 전에 실패를 다시 볼 수 있어야 한다.
우선순위:
규칙:
재현 불가 상태에서 해야 할 일:
관측 가능한 사실만 모은다.
항상 확인할 것:
멀티 컴포넌트 문제에서는 경계마다 확인한다.
예시:
각 경계에서 확인할 것:
문제 위치를 특정하기 전에는 고치지 않는다.
원인 후보를 하나만 세운다.
형식:
Hypothesis: <root cause> because <evidence>
좋은 가설의 조건:
나쁜 가설의 예:
원인을 소스까지 거슬러 올라간다. 오류가 깊은 스택에서 보이면 증상이 아니라 입력의 출처를 추적한다.
수정 전에 실패를 고정한다.
우선순위:
규칙:
자동화 테스트를 쓸 수 있으면 test-driven-development 스킬을 함께 사용한다.
수정은 하나의 가설만 다룬다.
허용:
금지:
실패하면 즉시 다시 Phase 1 또는 Phase 3으로 돌아간다. 이전 가설이 틀렸다는 뜻이다.
아래를 모두 만족해야 종료한다.
간헐 버그라면 한 번 통과로 끝내지 않는다. 반복 실행 또는 조건 변화 하 검증이 필요하다.
다음 상황이면 멈추고 프레임을 다시 잡는다.
여러 번 시도해도 재현이 안 되면:
재현이 안 되는데 코드를 바꾸는 것은 금지다.
세 번 연속으로 수정이 빗나가면 이렇게 판단한다.
이 시점부터는 "네 번째 땜질"이 아니라 구조 논의가 필요하다.
실패 테스트나 동등한 재현 장치를 만들 수 없으면, 완료로 선언하지 않는다. 최소한 재현 명령과 관측 결과를 남긴다.
아래 생각이 들면 즉시 멈추고 앞 단계로 돌아간다.
실행 중에는 아래 체크리스트를 기준으로 스스로 검증한다.
이 스킬의 완료 기준은 "코드가 바뀌었다"가 아니다.
완료 기준:
이 네 가지가 없으면 디버깅은 끝난 것이 아니다.
This skill includes reference documents for specific debugging techniques:
root-cause-tracing.md — Tracing bugs back through call chains to the original triggerdefense-in-depth.md — Adding validation at every layer to make bugs structurally impossiblecondition-based-waiting.md — Replacing arbitrary delays with condition-based pollingfind-polluter.sh — Bisection script for finding test pollution sourcescondition-based-waiting-example.ts — Complete implementation of condition-based waiting utilities장애 분석 및 운영 디버깅을 위한 확장 절차. 기본 Phase 1-7과 함께 사용한다.
phases/timeline-correlation.md — 배포/설정 변경 타임라인과 장애 시점 상관관계 추적phases/retry-cache-timeout-audit.md — retry/cache/timeout false-fix 안티패턴 체크리스트 (Hard Gate #7 구현)phases/amplification-detection.md — retry storm 및 error cascading 시그널 탐지phases/fault-injection.md — 가설 검증을 위한 의도적 장애 주입 절차| 상황 | 참조 |
|---|---|
| "언제부터 깨졌는가?" | phases/timeline-correlation.md |
| "retry/timeout 늘리면 되지 않나?" | phases/retry-cache-timeout-audit.md |
| 에러가 여러 서비스로 퍼졌다 | phases/amplification-detection.md |
| 가설은 있지만 재현이 안 된다 | phases/fault-injection.md |
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Pre-action boundary checking — validates agent tool calls against declared capabilities and task contracts
日本語の概要は準備中です。原文の説明を表示しています。
Auto-detect project context and optimize harness — deactivate unused agents/skills, suggest missing experts, generate project profile
日本語の概要は準備中です。原文の説明を表示しています。
Adversarial code review using attacker mindset — trust boundary, attack surface, business logic, and defense evaluation
日本語の概要は準備中です。原文の説明を表示しています。
Apache Airflow best practices for DAG authoring, testing, and production deployment
日本語の概要は準備中です。原文の説明を表示しています。
Alembic migration patterns for naming conventions, safety checks, expand-contract, env.py configuration, and CI integration
日本語の概要は準備中です。原文の説明を表示しています。
Pre-routing ambiguity analysis — scores request clarity and asks clarifying questions when needed (inspired by ouroboros)
日本語の概要は準備中です。原文の説明を表示しています。