Incidents

Incidents turn the catalog into a reliability tool. When something breaks, you declare an incident, attach the affected services, and Flagon routes it to the owning team and pages the on-call responder. It's the incident.io / Compass story, built on the services you already describe in the Catalog.

Declaring an incident

From Incidents, choose Declare incident (owner or admin). You give it:

  • A title and severity (SEV1 through SEV4).
  • Affected services: the catalog projects it impacts. The owning team auto-fills from the first affected service, so an incident always has an owner.
  • An optional escalation policy: who to page, and after how long, if nobody acknowledges (see On-call).

On declare, Flagon pages the first responder (the escalation policy's first level, or the owning team's on-call) by email, and attaches any matching runbooks so the recovery steps are already on the incident as a checklist.

The lifecycle

  • Timeline: post updates as you learn more; an update can change the status (open → investigating → identified → monitoring → resolved).
  • Acknowledge: any member can acknowledge, which stops escalation and signals someone has taken point.
  • Resolve: close it out. It drops off the open list and the affected projects' overviews.

Each project's overview shows its open incidents and who's on-call, so the catalog reflects live reliability state.

Over the API

Shell
# Declare
curl -X POST https://api.flagon.io/v1/orgs/{org}/incidents \
  -H "Authorization: Bearer $FLAGON_TOKEN" -H "Content-Type: application/json" \
  -d '{ "title": "Checkout failing", "severity": "sev1",
        "affectedProjectKeys": ["web"], "escalationPolicyKey": "sev1" }'

# Timeline + acknowledge + resolve
curl -X POST .../incidents/{number}/updates -d '{ "body": "Rolling back", "status": "identified" }'
curl -X POST .../incidents/{number}/acknowledge
curl -X POST .../incidents/{number}/resolve

Next

  • Runbooks: the playbooks that become an incident's checklist.
  • On-call: rotations, overrides, and escalation policies.
  • Catalog: the services incidents attach to.

Coming soon

A public status page (per-service status derived from open incidents) is the next piece of the reliability suite.