Agent
Adversarial audit of the retrospective's stage attributions and library health — verifies that each problem is attributed to its true upstream origin. Invoked in parallel with critic-coverage and critic-remediation at Submission.
Agent
Adversarial audit of the retrospective for exhaustiveness — verifies that every rule and knowledge source was enumerated, all six stages were examined, and gap analysis is complete. Invoked in parallel with critic-classification and critic-remediation at Submission.
Agent
Adversarial audit of the retrospective's remediation — verifies that each fix targets the stage where the cause lives (not a downstream patch), that implementable fixes were actually implemented, and that submission text follows style norms. Invoked in parallel with critic-coverage and critic-classification at…