error-recovery

Implement error recovery for transient failures using the retry with exponential backoff pattern, which automatically retries failed network requests or API timeouts with increasing delays between attempts to improve system reliability.

Handle CI/CD pipeline failures by applying structured error recovery patterns like diagnose-and-fix for build errors, graceful degradation for non-critical steps, and fallback alternatives to ensure deployment processes remain resilient to failures.

Graceful degradation works by allowing systems to continue operations when a non-critical step fails, delivering partial results instead of complete failure, ensuring microservices architecture maintains functionality during partial outages.

Escalate failures requiring human intervention by providing detailed context about the error, using the escalate with context pattern when retry logic and fallback alternatives are exhausted and the system cannot automatically recover.

Yes, you can use fallback alternatives to automatically switch to alternative tools or methods when the primary approach fails in your CI/CD pipeline, ensuring error recovery and continued operation without manual intervention.

Structured error handling for build, test, and linting errors enables squad agents to diagnose and fix issues with targeted recovery patterns rather than failing blindly, improving overall system reliability and agent coordination.