AgentAuth

This handles authentication and authorization primitives for AI agents that need to manage their own identity. You'd reach for this when building multi-agent systems where agents need to prove who they are, store and rotate credentials, delegate permissions to other agents, or enforce access controls on resources. Think of it as an identity layer for autonomous systems: instead of hardcoding API keys or sharing credentials across agents, you get proper token management and permission checking. The implementation details aren't extensively documented in the source, but the core value is clear if you're architecting agent workflows that require security boundaries or need agents to act with different privilege levels.