Post-Mortem & Incident Analysis

Top-Prompts

Audit failures. Document prevention. Maintain systems.

01 The Concept

When production outages occur, incident post-mortems must be direct, factual, and action-oriented. This template structures incident audits to extract timelines, discover root causes, and build concrete preventative tasks.

02 Weak vs. Strong

EX 01Database Connection Pool Outage

You are a Principal Reliability Engineer. Draft a post-mortem for yesterday's database outage.

Outage Details:

Incident: Connection pool exhaustion causing API 500 errors.
Timeline: Started 14:10 UTC; Detected 14:15 UTC via PagerDuty; Mitigated 14:35 UTC by doubling connection limit; Resolved 14:50 UTC after database restarted.
Root Cause: Unreleased connection in new /v1/orders endpoint route.

Format:

Executive Summary
Incident Timeline (UTC)
5 Whys Root Cause Analysis
Prevention Items (with owner/due date)

Tone: Blameless, analytical, technical.

→ Why it works

Structures the post-mortem into standard, actionable engineering sections. Ensures root-cause analysis is thorough rather than surface-level.

03 Key Points

01Timeline tracking: document event detection, mitigation steps, and complete resolution times.

025 Whys methodology: drill down through consecutive causes to locate the structural process failure.

03Prevention tasks: define exactly what checks, tests, or monitors will prevent recurrence.

04Objective tone: remove blame and emotional phrases. Focus on engineering gaps and telemetry.

04 Model-Specific Notes

Claude writes excellent blameless technical summaries. Highlight 'blameless culture' to align the tone.

05 For Your Role

A template that helps explain why a tech failure happened and lists what needs to be fixed to prevent it.

Next: Feature Spec & Implementation Plan →