analysis · Claude · free
Write a blameless post mortem for [INCIDENT] that a senior engineer would respect. Incident details: What broke, when, for whom: [DESCRIBE] Detection method + time: [WHEN + HOW] Mitigation + recovery time: [WHAT + HOW LONG] User impact (numbers): [SPECIFY] Rough timeline: [PASTE] Deliver: Summary — one paragraph, plain English Impact — quantified (users, minutes, $, SLA breach) Timeline — detection → response → mitigation → recovery, with UTC timestamps What went wrong — the causal chain, not "human error" What went right — the parts that saved us Where we got lucky 3 5 action items with owners and due dates (each testable) Root cause statement that names a system, not a person What we're changing so this class of incident is harder next time
#post-mortem #incidents #reliability