Santa Method is a multi-agent adversarial verification framework designed to catch systematic errors, biases, and hallucinations before outputs ship. The core insight is that a single agent reviewing its own output shares the same biases that produced it, so two independent reviewers with no shared context must evaluate the deliverable against a rigorous rubric. It features a structured four-phase process consisting of generation, independent dual review, a verdict gate, and a convergence loop for fixing issues. Reach for this skill when output will be deployed to end users, compliance or accuracy is paramount, and you need to ensure independent reviewers have both passed before shipping.