skill-security-auditor

To scan AI agent skills for vulnerabilities before installation, run a static code analysis on the skill directory using the security auditor script to detect malicious code and unauthorized access.

Static code analysis detects security threats in skill files such as command injection, data exfiltration, unauthorized access, and prompt injection in Markdown files.

Yes, you can use this security auditor to check Python and Bash scripts in skill directories, performing pattern matching and threat intelligence to identify potential security risks.

To run a pre-installation security check on a skill directory, execute the Python security auditor script and pass the target skill directory path as the argument.

Yes, prompt injection detection works on Markdown files by performing static analysis to identify malicious prompts and data exfiltration attempts before skill installation.

You need Python with subprocess, re, and json dependencies to perform pattern matching and threat intelligence operations during the static code analysis of skill files.