skill-vetter

To vet AI agent skills for security risks, you need a protocol that analyzes source code, permissions, and network access to detect malware and data exfiltration. The skill-vetter performs this analysis, classifying skills into risk levels from Low to Extreme.

Checking GitHub skills for malicious code requires analyzing the source for suspicious patterns and unauthorized access attempts. This security auditing protocol evaluates file and network scopes, assigning a risk classification to guide your installation decisions.

A security risk assessment for AI skills classifies permissions by evaluating the necessary file, network, and command scopes required by the skill. This scrutiny identifies unauthorized access attempts and assigns a risk level ranging from Low to Extreme.

Yes, you can use this vetting protocol to analyze skills from any source, including GitHub and ClawdHub. It ensures adherence to security best practices by checking for malicious code, data exfiltration, and unauthorized access before installation.

During code review, AI skills are assigned a risk level of Low, Medium, High, or Extreme. This classification guides installation decisions by indicating the severity of detected red flags, suspicious patterns, and unauthorized permission scopes.

You need malware detection for AI agent skills whenever installing a new skill from an unknown repository or source. Running a security review ensures the skill does not contain malicious code or attempt to steal sensitive information.