postmortem_skill

Overview

This skill helps teams run blameless postmortems to turn failures into learning opportunities. It guides you to capture a clear timeline, quantify impact, perform root cause analysis (5 Whys, fishbone, fault tree), and define SMART corrective actions with owners and deadlines. Use it to document lessons learned, prevent recurrence, and improve organizational processes.

How this skill works

The skill walks you through a five-step workflow: assemble a detailed timeline and impact metrics, conduct root cause analysis, define specific corrective and preventive actions, document and share the postmortem, and track action items to completion. It emphasizes blameless language, deep causal analysis, and actionability (specific owner and deadline for every item). Templates and prioritization guidance help you choose immediate fixes versus long-term investments.

When to use it

• After a production outage, service degradation, or security incident once the incident is resolved
• When a launch, project, or feature underperforms or misses goals
• For near-miss events that could have caused significant harm (proactive learning)
• When you need to understand systemic causes rather than assign individual blame
• During scheduled reviews (weekly/quarterly) to spot patterns and persistent pain points

Best practices

• Run the postmortem within 48 hours while memories are fresh
• Keep the process blameless: focus on systems and processes, not people
• Make every corrective action SMART: specific, measurable, assignable, realistic, and time-bound
• Quantify impact: users affected, duration, revenue/SLA impact, and relevant metrics
• Prioritize actions by impact and effort; track progress and close postmortems only when items complete

Example use cases

• Database outage: reconstruct timeline, estimate user impact and revenue loss, root cause via 5 Whys, implement config validation and staging
• Security breach: document detection-to-containment timeline, identify control gaps, assign remediation and audit tasks
• Failed product launch: analyze planning and testing gaps, create actions to improve discovery and stakeholder alignment
• Process breakdown: map steps, identify handoff failures, publish runbooks and checklists with owners
• Near-miss review: capture what almost went wrong and introduce preventive engineering controls