Tired of duct-taping access control into agent prompts. Here's the fix

This project introduces a structured, policy-based access control framework for AI agents, replacing the fragile and unscalable practice of embedding permissions directly into system prompts. It is designed for developers and security engineers building agentic systems who need granular, auditable, and maintainable authorization. The approach is interesting because it treats access control as a first-class architectural concern rather than a duct-taped workaround, enabling safer delegation of sensitive actions to autonomous agents.