Incidents

The Incidents page at /data-quality/incidents is the formal tracker for data-quality problems that need follow-up: a failed SLA, a confirmed anomaly, a contract violation. It plays the same role in DQ that PagerDuty plays for ops.

Lifecycle

open → acknowledged → investigating → resolved
                                    ↘ wont_fix

Incidents auto-open from:

• Anomalies — promoting an anomaly with Create Incident
• SLA Dashboard — failed SLA checks
• Data Contracts — contract validation violations
• Auto-Remediation — when an automated fix fails

You can also open one manually via New Incident.

What's tracked

Field	Notes
Title	Free-text
Severity	low, medium, high, critical
Status	Lifecycle state above
Affected tables	Multi-FQN
Owner	Email
Source	The detector that fired (anomaly / sla / contract / manual)
Description	Markdown-supported
Linked anomalies / runs	Click-through to the source detection

Filters

The header has filter pills:

• Status (open / ack / investigating / resolved / wont_fix)
• Severity
• Owner
• Date range

Timeline

Each incident has an activity timeline:

• Status changes
• Comments
• Links to remediation runs
• Auto-close events (e.g. anomaly resolved → incident auto-resolved)

API

GET  /data-quality/incidents?status=open
POST /data-quality/incidents
PATCH /data-quality/incidents/{id}        # update status / owner / fields
POST /data-quality/incidents/{id}/comments

Related

• Anomalies — primary auto-source
• SLA Dashboard — secondary auto-source
• Auto-Remediation — kick off a fix from an incident