Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope,...
Initial release of skill-vetter: a security-first vetting protocol for AI agent skills. - Provides a step-by-step guide for vetting skills before installation, focusing on source credibility, code review, permission scope, and risk classification. - Details red flags to watch for in skill code and mandates immediate rejection criteria. - Includes a standardized output report format for documenting vetting results. - Offers quick reference commands for vetting GitHub-hosted skills. - Establishes a trust hierarchy and best practices for skill installation decisions.