Skip to main content
Incidents are Tracelit’s unit of “something broke.” They are opened from frontend alerts or backend telemetry, enriched with context, and (when enabled) handed to AI Resolutions.
Whether incidents can open in a given environment is controlled in Utilities → Environment. See Utilities.

​
Where incidents come from

Detection only runs when the environment’s Allow incidents gate is on. Local backend traffic often keeps incidents off by design.

​
Environments

Incidents are scoped to the environment that produced the signal. Archiving a site or service stops new incidents for all of its environments.

​
Incident lifecycle

1

Detected

Tracelit fingerprints the error or anomaly. New fingerprints (or spikes above baseline) open an incident.
2

Open

The incident appears in the Incidents list with severity, summary, and linked evidence (replay for sites; logs / traces for services).
3

Investigating

Your team reviews the timeline, attributes, and linked telemetry. Notifications can include Slack when connected.
4

Resolution (optional)

If resolutions are enabled for that environment and a repository is linked, you can start an AI Resolution from the incident.
5

Resolved or ignored

Mark resolved when fixed, or ignore known noise. Ignored fingerprints will not keep reopening as active work.

​
What an incident includes

​
Enabling incidents

1

Open Utilities

Select the site or service in the app rail, then open Utilities.
2

Environment tab

For each environment, enable Allow Incidents Run.
3

Save

Gates apply immediately to new telemetry. Existing open incidents are unchanged.
If incidents never appear, check the environment gate first — then confirm ingest is reaching Tracelit for that environment.

​
Notifications

Connect Slack under Settings → Integrations. Incident cards can include severity, summary, and a link back to the dashboard.

​
Next