Bindu wrestles with agent trust controls

The missing primitive is verifiable delegation, not another human-style login flow. Bindu is already close to the right abstraction, but the policy has to travel with the message or every worker becomes a potential confused deputy.

• –Bindu's docs already expose trust metadata like `agentTrust`, role inheritance, verification requirements, and `allowedOperations`, which is the right shape for machine-to-machine auth.
• –DID identity and encrypted transport establish who is speaking, but not what that agent is allowed to do on behalf of others.
• –Auth0/Ory Hydra and JWTs can cover service authentication; delegation still needs scoped, signed capability claims that other agents can verify.
• –For multi-framework swarms, trust has to live at the protocol layer, not just inside one SDK, or every integration will reinvent policy wrappers.
• –If Bindu gets this right, it becomes closer to IAM for agents than just another orchestration layer.