Security audit skill for AI agent skill packages — detects credential exfiltration, prompt injection, and undisclosed behavior before install
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).