Santa Method

Two independent review agents evaluate your output against a rubric before it ships. Neither sees the other's assessment, breaking the bias problem where a single reviewer shares the same blind spots as the generator. Both must pass or the output loops back through a fix cycle, up to three iterations before human escalation. The rubric is everything here: vague criteria get vague reviews, so the framework pushes you to define objective pass/fail conditions for accuracy, completeness, compliance, whatever matters for your domain. Best for production code, customer-facing content, or anything where hallucinations and systematic errors are expensive. Overkill for drafts and exploration.

Install to Claude Code

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