Skip to content
Parampt
Start free
Open menu

Development · Free interactive template

Incident Postmortem Prompt for ChatGPT and Claude

Turn an incident timeline and evidence into a blameless postmortem with contributing causes and verifiable actions. Free to try without an account.

No account or credit card needed to prepare and copy your prompt.

Make this prompt yours

Free to prepare and copy. Your AI tool produces the response.

Your completed prompt

Fill in the form, or load the fictional example. Your instructions will appear here, ready to paste into the AI tool you choose.

Keep it for next time

Save this template with your inputs, then edit and reuse it in your library. Your first 5 saved prompts are free.

When you save, a draft stays in this browser for up to 7 days to continue after sign-in. Use the same browser and device. Reset clears it.

When to use this template

  • Writing the postmortem draft the day after an outage, while the details are fresh
  • Turning a messy incident channel into a timeline the whole team agrees on
  • Preparing a short summary of a technical incident for nontechnical stakeholders

How to get a useful result

  1. Paste timestamps exactly as they appear in your alerts and chat. Approximate times hide the gap between detection and diagnosis.
  2. Write down the hypotheses the team ruled out. They explain why the response took the time it did.
  3. Remove customer data, credentials, and internal hostnames from logs before pasting them into an AI tool.

What this template does—and its limits

The AI only knows the evidence you paste. It cannot inspect your systems or confirm a root cause, so have the people involved review the draft before it is shared.

Read the full reusable prompt
You are a Site Reliability Engineering lead. Write a blameless postmortem for this incident: {{incident}}.

Timeline: {{timeline}}
Impact: {{impact}}
Technical evidence: {{evidence}}
Response: {{response}}

Establish a factual timeline distinguishing event time, observation time, and decision time. Quantify only demonstrated impact. Explain the technical failure mechanism, then contributing conditions involving the system, processes, interfaces, and assumptions. Avoid “human error” and simplistic single-cause narratives.

Analyze detection, diagnosis, coordination, communication, mitigation, and recovery: what helped, slowed things down, or surprised the team. Propose actions across prevention, blast-radius reduction, detection, and response, each with priority, responsible role, verifiable outcome, and a deadline to define if missing. Finish with lessons, residual risks, and a short version suitable for nontechnical stakeholders.

More prompt templates

Browse all public templates