
from meta-skills12
An independent reviewer-resolver loop that checks code or artifacts against requirements without sunk-cost bias to ensure high quality before shipping.
This skill implements a professional quality assurance loop. It spawns a 'Reviewer' agent who has no access to the implementation reasoning—only the output and requirements—to identify flaws. A 'Resolver' then synthesizes these findings to fix the work. This process prevents 'sunk-cost bias' where the original author overlooks their own mistakes.
Trigger this skill for security-sensitive code (auth, crypto), data mutations (migrations, bulk deletes), complex implementations, or any critical artifact that needs a second opinion before a production push.
Designed for advanced agentic environments including Claude Code, Cursor, Codex, Windsurf, and Gemini CLI that support multi-agent spawning or complex state management.
This skill has not been reviewed by our automated audit pipeline yet.