Security scanning for AI agent skills exposed as MCP tools, enabling skill analysis from ClawHub, GitHub, skills.sh, or raw URLs.
