Detect leaked API keys and credentials in text or URLs
Detect malicious patterns in AI agent skills and prompts