skill-security-auditor

To scan AI agent skills for security vulnerabilities before installation, you can run an auditing tool against the local skill directory to analyze Python and Bash scripts for malicious code, providing a PASS/WARN/FAIL verdict with remediation guidance.

Prompt injection detection in Markdown files is the process of analyzing text content to identify malicious instructions hidden in documentation that could manipulate an AI agent's behavior, preventing unauthorized command execution during skill integration.

Yes, you can audit Python and Bash scripts for command execution risks by scanning the skill directory to detect dangerous code patterns like command injection and code execution vulnerabilities, returning a clear security verdict.

You check dependency supply chain integrity for untrusted AI skills by running a pre-installation security audit that analyzes external dependencies for potential supply chain issues, ensuring community-contributed skills are safe to integrate.

Yes, this vulnerability scanner works on local skill directories by analyzing the specified local path to detect execution risks, prompt injection, and dependency issues, outputting a clear PASS, WARN, or FAIL verdict.

A limitation of pre-installation security scanning for AI skills is that it analyzes static code patterns and local dependencies, meaning dynamically loaded payloads or runtime environment exploits might not be detected before installation.