Postmortems

The incident is resolved, everyone's asleep again, and the timeline is still warm. Postmortems turn that timeline into a written record — generated from the incident's audit trail, editable where the machines get it wrong, and exportable for the wiki or the PDF archive.

Before you start
  • An incident with some history — per Incidents & services. The postmortem is assembled from the incident's audit trail, so a blank incident makes a thin report.
  • Admin or team access to the incident. Anyone who can read the incident can generate and read its postmortem.

  1. Generate from the incident

    Open the incident detail page and switch to the Postmortem tab, then hit Generate. OpsPing assembles the report from the incident's audit trail: when it started and ended, who responded and when, status updates, priority changes, and linked alerts. No copy-paste from the timeline, no reconstructing timestamps from memory three days later.

    Generate works best after the incident is resolved — then the whole trail exists and the report can state real durations instead of "still ongoing."

  2. Edit what the machine can't know

    The generated report gets the facts right; it can't know why the deploy took the database down. Fill in the sections that need a human: the summary, the root cause, and the action items. Action items are the part that actually matters — the concrete fixes that stop this incident from happening again.

    Missed something in the timeline and need to regenerate? Fine. Regenerating rebuilds the audit-trail sections but preserves your edits — your summary, root cause, and action items survive the rebuild.

    Tip

    Write the postmortem within a day or two of resolving, while the details are fresh. A postmortem written from a fading memory is how the same incident happens twice.

  3. Export to Markdown or PDF

    When the report is ready, export it. Markdown drops straight into a repo or wiki; PDF is for the archive, auditors, or anyone who asks for "the report." Both exports reflect the current state of the postmortem, edits included.

    Use the export buttons on the postmortem view, or fetch the file via the API:

    curl "https://api.ops-ping.com/v2/postmortems/POSTMORTEM_ID/export?format=pdf" \
  -H "Authorization: OpsPingKey YOUR_API_KEY" \
  --output postmortem.pdf
    Note

    Swap format=pdf for format=md to get the Markdown export instead.

  4. Do it from the API

    Everything above is scriptable. Generate a postmortem for an incident:

    curl -X POST https://api.ops-ping.com/v2/postmortems/generate \
  -H "Authorization: OpsPingKey YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"incidentId": "INCIDENT_ID"}'

    The response includes the postmortem's id, which you use for the export call above. See the API reference for the full request and response shapes.

Next up Status pages →