메뉴
Root-cause analysis for bugs, incidents, and regressions — timeline, hypothesis testing with evidence, 5-whys, prevention rules feeding the memory loop. Use when investigating a production bug, an error log the user dropped in, or a recurring defect.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Read and understand an unfamiliar/legacy codebase safely before changing it — entry points, dependency tracing, behavior mapping, characterization tests, seams. Use during onboarding of an existing project or before any change in code you did not write.
Angular v19-v21 features: linkedSignal, resource/httpResource, signal forms, route-level SSR render modes, incremental hydration, zoneless change detection, Vitest migration, and Angular Aria components.
Master REST and GraphQL API design principles to build intuitive, scalable, and maintainable APIs that delight developers. Use when designing new APIs, reviewing API specifications, or establishing API design standards.
Astro changes since training cutoff (latest: 6.0) — Fonts API, live collections, CSP, route caching, ClientRouter, Zod 4/Vite 7/Shiki 4. Load before working with Astro.
Axum changes since training cutoff (latest: 0.8.5) — {param} path syntax, no async_trait, OptionalFromRequestParts, Utf8Bytes WS, HTTP/2 WebSockets, NoContent. Load before working with Axum.
Guidelines for Blazor development including component lifecycle, state management, and performance optimization
| name | incident-rca |
| description | Root-cause analysis for bugs, incidents, and regressions — timeline, hypothesis testing with evidence, 5-whys, prevention rules feeding the memory loop. Use when investigating a production bug, an error log the user dropped in, or a recurring defect. |
| category | meta-process |
Turn "it broke" into an evidence-backed cause and a prevention rule. Pairs with bug-router (R-009) and the feedback loop (R-010-09).
1. REPRODUCE or capture: exact error, inputs, environment, frequency. No repro -> say so explicitly
and work from logs/timeline with lower confidence (state it).
2. TIMELINE: what changed before it broke? (git log on the touching area, deploys, config, data,
dependency bumps). Correlation candidates, not conclusions.
3. HYPOTHESES: list 2-4, each with a CHEAP TEST that would confirm/kill it. Test in cheapest-first
order; record each result (journal, R-024-03).
4. ROOT CAUSE: apply 5-whys until you reach a process/design cause, not just the faulty line.
"Null pointer" is a symptom; "contract change unannounced + no contract test" is a cause.
5. FIX + REGRESSION TEST: the fix ships WITH a test that fails before / passes after (evidence).
6. PREVENTION: write root_cause, prevention_rule, regression_check, recurrence_key into
coding_error_feedback / feedback memory (R-006-17, R-010). Recurring pattern (>=3) -> draft a
proposed rule (R-010-09).
Blocker vs non-blocker per R-009 — never downgrade to pass QC (R-019-QC4).
Distinguish: defect (code wrong) / regression (was right, broke) / config / data / external.
The category changes who fixes it and where the prevention lands.
No "probably fixed": done requires the regression test passing (R-023-03 evidence).
If two causes remain plausible, say which one you fixed and why the other is ruled out (or not).