Agent Policy

A behavioral governance mechanism for AI agents that defines the set of permitted and prohibited operations an agent may execute. It encompasses three core dimensions: intent recognition, role-based access control, and approval workflows, ensuring agents operate within defined safety boundaries.