IT incidents, resolved and proven fixed.

Rook is an operations agent for growing companies. It investigates across your systems, fixes in a safe order, recovers when a step fails, and closes an incident only after independent checks pass.

Incident type

  1. 01InvestigateRook reads every connected system and maps what is affected. Nothing is changed yet.
  2. 02PlanIt writes the fix as ordered steps, plus the checks that must pass before it can call the incident closed.
  3. 03Fix safelyEach step is checked against company rules. Risky ones wait for a person. Failures are inspected, not blindly retried.
  4. 04Prove itIndependent checks confirm the fix really worked. If they fail, Rook says so and hands over to a person.
Start
Pick an incident type above and click Investigate.

Access & offboarding: someone left, but their access lives on. Service outage: a live service broke after a deploy. Both run on fake systems, so nothing real is touched.

sandboxmock integrationsorg …loading

Incidents

Each run is stored with its plan, checkpoints and audit trail.

Loading…

    Try one of two incidents

    Access & offboarding: a contractor left but their API key still runs two services. A config API fails once mid-fix; Rook checkpoints, inspects what really happened, retries safely and closes with evidence.

    Service outage: checkout-api broke right after a deploy. Rook pauses the pipeline so it can’t redeploy the bad change, rolls back, confirms customers can check out, and turns the pipeline back on with the bad revision blocked.