Ravi Kiran Kadaboina

The paper proposes Anumati, a formal consent model that moves beyond simple proof of acceptance to provide a verifiable, per-action proof of adherence to evolving policies in autonomous agent communication.

Pramana introduces a standardized, protocol-level wire format for autonomous agent outputs, ensuring that every consequential claim is accompanied by a verifiable artifact that can be re-executed by an external auditor.