概要
The Validation skill provides a robust framework for auditing AI development plans before execution, preventing costly errors in complex codebases. It employs a two-tier approach: an adversarial 'Challenge Review' that rigorously tests plans against edge cases, simplicity, and integration requirements, followed by a 'Confidence Scoring' system that quantifies readiness across four critical dimensions. This ensures that every proposed change is atomic, verifiable, and aligned with existing architectural patterns before a single line of code is written, effectively acting as a pre-execution quality gate.
