Repair workflow
Investigate an incident
Read the timeline and identify the failing contract.
Before you begin
- A Seamward incident
- Access to the affected workspace
- Representative observation and finding evidence
Review the summary
Confirm the affected integration and incident window.
Confirm the incident status, integration, first and latest occurrence, and affected observation count before drawing conclusions from an individual event.
Inspect evidence
Compare findings with linked observations.
Read the timeline in order. Compare the finding path, previous value type, observed value type, and linked observation before deciding that a provider change caused the incident.
Check source context
Review the release and commit associated with the failure.
Use the observed release and commit SHA to identify the application version that processed the failing traffic. Repository access is optional for investigation but required for GitHub source delivery.