saga-ed

Set up a QA team with two independent testers that cross-validate each other's work. Uses Claude Code experimental agent teams to run parallel test passes, swap sections, and produce an agreement matrix highlighting confirmed passes, confirmed failures, and disagreements requiring investigation.

Systematically evaluate repository compliance with documented constraints in hierarchical CLAUDE.md files. Generates compliance reports with 5-point ratings for each app/package across soa, coach, and thrive repositories. Use when you need to assess codebase health, verify adherence to documented patterns, or identify non-compliant packages.