PromptSharp › Prompt Library › Dev & Engineering › Blameless postmortem draft: timeline, contributing factors, and actions that stick

Blameless postmortem draft: timeline, contributing factors, and actions that stick

An incident is resolved and the postmortem is due. Draft a blameless write-up from the chat logs and alerts while the details are fresh.

The prompt — copy and run it

You are a site-reliability engineer drafting a blameless incident postmortem. I will paste the alert history, the incident chat log, and the fix that was deployed.

Produce:

A) SUMMARY — what happened, user impact, duration, and how it was detected, in 4 sentences.

B) TIMELINE — timestamped events from my paste: first signal, detection, escalation, key decisions, mitigation, resolution. Mark gaps where the logs are silent.

C) CONTRIBUTING FACTORS — technical and process factors, each phrased as a condition in the system, not a person's mistake.

D) WHAT WENT WELL AND WHAT SLOWED US — detection, communication, tooling, runbooks.

E) ACTION ITEMS — each one specific, owned by a role, dated, and tagged prevent, detect faster, or recover faster. Limit to the 5 with the most impact.

Inputs: [PASTE ALERT HISTORY] · [PASTE INCIDENT CHAT LOG] · [THE FIX DEPLOYED] · [USER IMPACT DATA]

Rules: Do not invent timestamps, causes, or impact numbers — use only my paste and mark anything inferred. Keep it blameless: refer to roles, never named individuals. Verify the timeline against the raw logs before publishing. Strip customer data, credentials, and personal identifiers from anything quoted.

Why this prompt works

Postmortems fail when they blame a person, which teaches people to hide mistakes, or when they end in vague actions nobody owns. Framing contributing factors as system conditions, building a timeline only from logs, and limiting actions to five owned, dated items tagged by type produces fixes that actually happen. Marking silent gaps shows where monitoring is missing.

Frequently asked

Related Dev & Engineering prompts