Skip to content
Opsrift
Opsrift · Incident documentation, automated

Nobody has three hours
for a postmortem.

So Opsrift writes the first draft. Nine sections from your PagerDuty incident, ready in under a minute. You review what matters and publish to Confluence.

Generate a postmortemTry it without an account
7-day trial · cancel anytime
PagerDutyOpsGenieDatadogGrafanaJiraConfluenceSlackMicrosoft TeamsGitHub
By hand~3h
  1. Open a doc. Stare at the blank template.
  2. Rebuild the timeline from Slack scrollback.
  3. Pull commits, alerts, and graphs from six tabs.
  4. Ping three people for the parts only they know.
  5. Format it. Reformat it. Fix the table again.
  6. Publish, then do all of it again next incident.
With Opsrift47s
  1. Pick the incident from PagerDuty or Jira.
  2. Nine sections draft in parallel.
  3. Fix anything off. Regenerate a section if needed.
  4. Action items push to Jira as tickets.
  5. Publish to Confluence in one click.
  6. Walk into the retro with it already done.
The platform

Six things you stop
writing by hand.

  1. 01Postmortem GeneratorNine sections drafted from one incident: summary, timeline, root cause, impact, response, resolution, gaps, action items, lessons. Import it, review it, publish to Confluence.
  2. 02Shift HandoverSeven sections covering the window you worked: what is open, what is pending, what is escalated, what is coming. The next on-call reads it instead of joining a call.
  3. 03Runbook GeneratorEight-section runbooks from scratch, from an alert, or straight from a postmortem you already wrote. Diagnostics, fixes, rollback steps, escalation contacts.
  4. 04Incident InvestigationAsk what happened in plain English. Answers stay grounded in your PagerDuty, Datadog, Grafana, and Jira data, and hold context across follow-ups.
  5. 05Status PageA hosted status page on your own domain. Component health, incident updates, scheduled maintenance, all public-facing.
  6. 06Incident ForecastReads your postmortem history for what is coming: which systems keep breaking, which deploys correlate with incidents, how much action-item debt is open.
Integrations

It reads from the stack
you already run.

PagerDutyImport incidentsOpsGenieImport alertsDatadogImport alerts and tracesGrafanaImport panels and alertsJiraImport issues, push action items backConfluencePublish finished docs in one clickSlackPost incident summaries to a channelMicrosoft TeamsPost incident summaries to a channelGitHubCorrelate deploys with incidents
Most incidents never get a postmortem. Not because the team doesn’t care. Because the incident ends at 3am, and nobody has another three hours left in them. Opsrift takes the three hours out, so the writeup actually happens.
Why we built Opsrift
Questions

The short answers.

What does Opsrift actually do?
It writes the documents on-call teams produce after and around incidents. Postmortems, shift handovers, and runbooks are the core. Incident investigation, a hosted status page, and incident forecasting round it out. Six tools, one platform, all live today.
How does it write a postmortem?
Pull an incident from PagerDuty, OpsGenie, Datadog, Grafana, or Jira, or enter the details by hand. Claude drafts nine sections in parallel: executive summary, timeline, root cause, impact, detection and response, resolution, process gaps, action items, and lessons learned. You review and edit before any of it ships.
Which tools does it connect to?
PagerDuty and OpsGenie for incidents. Datadog and Grafana for alerts and metrics. Jira for issue import and action-item push. Confluence for one-click publishing. Slack and Microsoft Teams for notifications. GitHub for deploy correlation. All nine are available now.
Can I fix what the AI gets wrong?
Yes, and you should. Every section edits inline, and you can regenerate any single section without touching the rest. The draft is a starting point, not a verdict. Nothing publishes until you decide it is right.
What is in the Starter plan?
Ten generations a month, every core tool, one status page, PagerDuty and OpsGenie import, and exports to PDF, Markdown, and Slack. It is $10 a month. Pro drops the generation limit for $29 a month. Both start with a 7-day trial.
How do shift handovers work?
Connect PagerDuty or Jira, set your shift window or use the 8h, 12h, and 24h presets, and Opsrift builds a seven-section handover: summary, open incidents, pending tasks, escalations, metrics, upcoming changes, and notes. The next on-call reads it instead of sitting through a passdown call.
What is Incident Forecast?
It reads your postmortem history for what is likely coming. Systems that keep recurring, deploys that correlate with incidents, unresolved action-item debt, severity trend, and mean time between incidents. A short risk brief sums it up so you can act before the next P1, not after.
Where does my incident data go?
Into your account, and nowhere else. Authentication runs on Clerk, storage on Supabase with row-level security, and every request is encrypted in transit. Your data is used to generate your reports. It is never used to train AI models.

The next incident is coming.
Be ready to write it up.

Start a 7-day trial. The postmortem generator runs without an account, so you can test it on a real incident before you connect a single tool.

Start the trialSee pricing