Automated AI Security & Vulnerability Engine
AI Skill & MCP Security Checker
Scan any GitHub repository, Model Context Protocol server, or SKILL.md package for dependency vulnerabilities, prompt injection exploits, and credential leaks before installing.
Try popular:••
Automated CVE Audits
Every indexed skill is continuously tested against GitHub Advisory databases, pip-audit, and npm-audit for known vulnerabilities.
Prompt Injection Defense
Heuristic parsers scan system prompts and tool schemas to prevent malicious jailbreaks and indirect role takeover attacks.
AST Sandboxing
Validates that terminal skills execute within local sandboxes and do not leak environment variables or private API tokens.