Protects agent skills from being distilled during multi-agent collaboration by automatically filtering sensitive content, with features for skill declaration, output filtering, audit logging, and license management.
