Incident Post-Mortem Generator: Turn Outages Into Playbooks

When something breaks at 3 AM, you fix it. But without a post-mortem, the same thing breaks again next week. This skill generates structured, actionable incident post-mortems automatically — pulling timeline data from PM2 or systemd logs, identifying root causes, and producing a complete document with timeline, impact assessment, action items, and preventive automations.

Designed for AI agents and SRE teams running production services. Includes severity classification, log extraction scripts, auto-detection of PM2 restarts, and a framework for turning action items into actual code (monitoring scripts, restart policies, cron checks) — not just documents that gather dust.

Covers the full lifecycle: incident detection → timeline reconstruction → root cause analysis → post-mortem document → preventive automation. Built from real production incident patterns.