Incident Postmortem Generator

Turn incident notes into a postmortem with summary, impact, timeline, detect/mitigate/resolve times, root cause and owned action items, as Markdown.

The interactive tool loads here when JavaScript runs.

About this tool

What is the Incident Postmortem Generator?

The Incident Postmortem Generator turns your incident notes into a structured review. Enter the title, date, severity, summary and impact, paste the timeline, describe the root cause, the response and the lessons learned, and record time to detect, mitigate and resolve in minutes. Add action items with an owner, priority, due date and category (detection, mitigation, prevention or solution). Generate builds the report, which you can copy or download as Markdown.

The report only contains what you entered. Timeline lines that start with a time ("14:30 - Alert fired", "2026-03-02 14:30 UTC: paged") are split into a time and an event, minutes are shown as "1 h 30 min", and any field you leave blank appears as "Not recorded" instead of a made-up number. A review with no action items says so, because that is usually the gap worth fixing.

Nothing is uploaded or stored; download the Markdown before closing the page.

  • Structured form with a complete sample incident
  • Timeline parsing for common time formats
  • Detect, mitigate and resolve durations
  • Action items with owner, priority, due date and category
  • Markdown export; nothing invented, nothing stored
Read the guide: Blameless Incident Postmortems: A Template and How to Run One

Updated

How to use the Incident Postmortem Generator

  1. Fill in Incident Title, Incident Summary and Impact Assessment (required), plus the date and severity. Load Sample shows a complete example.
  2. Paste the Timeline of Events with one event per line, starting each line with a time where you have one.
  3. Describe the Root Cause, Response and Resolution, and Lessons Learned (one per line). Enter Time to Detect, Mitigate and Resolve in minutes, and the Detection Method.
  4. Add action items: title, description and owner are required; set priority, category and due date, then press the add button.
  5. Press Generate Incident Report, then Copy to Clipboard or Download Report.

Worked examples

Timeline parsing

Input

09:14 - Alert fired for 5xx rate
09:20 - Incident declared
Rolled back release 412

Output

- **09:14** Alert fired for 5xx rate
- **09:20** Incident declared
- Rolled back release 412

Lines without a timestamp are kept in order as events.

Durations

Input

Time to Mitigate: 25, Time to Resolve: 90, Time to Detect: (blank)

Output

Time to mitigate: 25 min, Time to resolve: 1 h 30 min, Time to detect: Not recorded

A blank metric stays blank in the report, so nobody reads a guess as a measurement.

Background and insight

The blameless postmortem became mainstream practice through web operations teams around 2010 to 2012, notably Etsy's writing on it, and Google's Site Reliability Engineering book (2016) gave it a full chapter. The idea is simple: people act reasonably on the information they had, so a review that hunts for someone to blame teaches everyone to hide information next time. Write "the deploy script did not check for X", not "Sam forgot to check X".

Three durations make reviews comparable over time. Time to detect runs from the start of impact to the first alert or report. Time to mitigate runs to the point users stop being hurt, often through a rollback or failover rather than a fix. Time to resolve runs to the real fix. Teams often improve mitigation far faster than resolution, which is fine: users care about the first two.

Common use cases

  • Writing the review after an outage

    Turn the incident channel notes into a readable document for the review meeting.

  • Standardizing reviews across teams

    Give every team the same sections so reports can be compared month to month.

  • Tracking follow-ups

    Collect owned, dated action items grouped by whether they improve detection, mitigation or prevention.

  • Customer-facing summaries

    Start from the executive summary and impact sections when drafting a status page update.

Tips and limitations

  • Write the timeline in UTC and say so; mixed time zones are the most common timeline error.
  • Separate the trigger (a config change) from the root cause (nothing validated pool size changes). Fixing only the trigger invites a repeat.
  • Every action item needs one owner, not a team alias, and a date. Unowned items do not get done.
  • Avoid names in the root cause section; describe the system and process gaps instead.
  • Hold the review within a week, while logs and memories are fresh.

Sources and further reading

  1. Postmortem Culture: Learning from Failure, Google. Chapter 15 of Site Reliability Engineering
  2. Postmortem Practices for Incident Management, Google. Worked examples of good and bad postmortems
  3. Postmortem Process, PagerDuty. When to write the review, who owns it and what it covers

Frequently asked questions

What should an incident postmortem include?

A summary, the customer impact, a timestamped timeline, how the incident was detected, the root cause and contributing factors, what went well and badly in the response, and action items with owners and dates. This generator has a section for each.

What is a blameless postmortem?

A review that treats the incident as a failure of systems and processes rather than of individuals. It assumes people made reasonable decisions with the information they had and asks what made the mistake easy, so engineers keep reporting problems honestly.

What is the difference between time to mitigate and time to resolve?

Time to mitigate ends when users stop being affected, for example after a rollback. Time to resolve ends when the underlying problem is actually fixed. Mitigation often takes minutes while resolution can take days, and both are worth recording.

How soon after an incident should we write the postmortem?

Draft it within a few days and hold the review within a week. Logs rotate, chat history gets buried and people forget the order of events. Capture the timeline during the incident if you can.

Does the incident review tool save my reports?

No. The report exists only in your browser tab. Copy it or download the Markdown file, then store it wherever your team keeps incident records, such as a wiki, repository or incident tracker.

Is there an in-depth guide to the Incident Postmortem Generator?

Yes. “Blameless Incident Postmortems: A Template and How to Run One” covers the background, how it works, common pitfalls and worked examples. Read the full guide.