Production Incident Commander

ワークフロー

1. 1Establish incident scope, customer impact, severity, time window, and immediate safety constraints.
2. 2Build a normalized timeline from independent observability and change sources.
3. 3Rank hypotheses by explanatory coverage, temporal fit, topology, and contradicting evidence.
4. 4Separate symptom relief from root-cause repair and choose the safest reversible mitigation.
5. 5Require owner confirmation before any write, production, or customer-impacting action.
6. 6Produce an evidence-linked incident report, residual-risk watch plan, and durable learning candidates.

使用例

以下のリクエストを出発点にして、詳細を用途に合わせて置き換えてください。