/circuit-breaker

To handle repeated validation failures, you need systematic course correction to analyze failure context, assess severity, discard disposable code, and guide recovery through spec modification. This prevents looping over the same implementation errors.

The best way to rollback a design during critical issues is to initiate a course correction that assesses whether the failure is local or upstream, discards flawed generated code, and preserves the spec-as-source principle for re-execution.

You can determine if a recurring bug is local or upstream by using automated failure analysis to summarize the failure context, which differentiates between minor implementation issues and major architectural flaws to guide the recovery approach.

Yes, you can modify specifications instead of discarding all code. Course correction ensures the spec-as-source principle by preserving specifications over generated code, clearly defining which artifacts to discard and which to keep during recovery.

You should use a design rollback instead of fixing the implementation directly when severity assessment categorizes the validation failure as a major architectural flaw rather than a minor implementation issue, requiring deeper design correction.