error-recovery

Workflow failure recovery is automated by classifying the failure cause and applying auto-fix strategies with up to three retry attempts. If the failure classification determines manual intervention is required, it escalates with structured remediation guidance.

Classifying CI pipeline failures involves using a failure classification tree to categorize the error and determine auto-recoverability. The system evaluates the failure state to decide if safe automatic retries are possible or if manual remediation is needed.

Auto-recovery handles blocked deployments by identifying the quality gate failure and applying targeted automatic recovery strategies. It caps these efforts at three retry attempts to prevent infinite loops before escalating the issue for manual intervention.

Use an automated error-recovery strategy when sprint or feature implementations fail and require immediate classification. It determines if the workflow phase failure is safely auto-recoverable, logging outcomes and escalation paths for any unresolved manual interventions.

Automated retry logic does not work for all build and test failures, as it depends on the failure classification tree. It only applies auto-fix strategies for safely recoverable categories, capping at three attempts before transitioning to manual intervention.

Limitations of automatic recovery include a strict cap of three retry attempts for safe auto-fix strategies. Failures classified outside the auto-recoverable categories immediately require manual intervention and structured escalation prompts instead of automated retries.