Santa Method

Two independent agents review your output against the same rubric with zero shared context, and both have to pass before anything ships. The insight is simple: one agent reviewing its own work carries the same blind spots that created the output in the first place. Built for production code, customer-facing content, and compliance-sensitive work where hallucinations or systematic errors matter. If either reviewer flags an issue, a fix agent patches it and both reviewers run again on fresh instances with no memory of prior rounds. Caps at three iterations before escalating to human review. Ronald Skelton built this at RapportScore.ai for catching the stuff that slips through when you're the only one looking at your own output.

Install to Claude Code

npx -y skills add affaan-m/everything-claude-code --skill santa-method --agent claude-code