skill-vetter

To vet AI skills for security risks before installation, perform source verification, mandatory line-by-line code review, permission scope analysis, and risk classification to identify data exfiltration or unsafe behavior.

Permission scope analysis for code review evaluates read, write, and file access along with network calls and runtime privileges to identify potential data exfiltration or unsafe behavior in AI skills before deployment.

Assess author reputation and source of GitHub or ClawdHub skills by verifying origin, checking update history, and performing a mandatory code review to detect risky code prior to integration in trusted environments.

Yes, you can generate a risk assessment report for AI skills across deployment pipelines by categorizing risk levels and producing a structured vetting report for decision makers after completing source verification and code review.

Skill vetting works for unknown sources in development environments by applying a multi-step vetting protocol that verifies origin, analyzes permission scope, and classifies risks before installation.

Risk classification for AI skills generates a structured report for decision makers but should be combined with mandatory line-by-line code review, as automated classification alone may miss nuanced red flags or complex data exfiltration techniques.