← All use cases

Engineering

Run on demand

The postmortem gets a first draft.

A workflow rebuilds the timeline from your incident channel and the deploys around it. It names the likely change and drafts the write-up with the facts in place.

5,000 free credits to start.

How it works

Three steps to a workflow

  1. Step 01

    Point it at the incident channel

    Give the agent the Slack channel and the repositories involved. It reads the whole conversation, including threads, and pulls the commits and deploys inside the impact window.

  2. Step 02

    Run it when the incident closes

    Start the run by hand from the workflow page, or from chat with the channel name. A schedule can also sweep for channels created in the last day.

  3. Step 03

    Start from a draft

    The postmortem lands in Notion with the timeline, the contributing factors, and the action items people agreed to in the channel.

The run

What one run looks like

Each line is a tool call. The agent finds the API it needs, reads the result, and acts. The full transcript stays on the run.

Run on demand

Draft a postmortem from the #incident channel and the deploys around it.

  1. slack·read #incident-2026-03-06 in full
  2. github·list deploys and reverts in the window
  3. github·read the diff for the suspect deploy
  4. code·build the minute-by-minute timeline
  5. notion·create the postmortem draft

Postmortem drafted

  • 41-minute impact window rebuilt
  • Likely deploy identified as #2833
  • Detection took 12 minutes, alerting gap noted
  • 5 action items pulled from the discussion

Integrations

The tools this workflow uses

Connect each one once. The agent gets its full API, and the credential stays at the network layer.

FAQ

Questions, answered

Put a workflow on incident summaries.

Write the prompt, pick the trigger, and read the first run.