Anumati: Proof of Adherence as a Formal Consent Model for Autonomous Agent Protocols
For developers and users of autonomous AI agent systems, this work provides a formal mechanism to ensure agents can prove they understood and honored terms of service, addressing a structural accountability gap.
The paper addresses the accountability gap in autonomous agent-to-agent communication by formalizing a consent model that distinguishes proof of acceptance from proof of adherence. It proposes three primitives (PolicyDocument, ConsentRecord, AdherenceEvent) as a non-breaking extension to A2A and MCP protocols, with a TLA+ specification and Python implementation.
As autonomous AI agents increasingly call other agents to complete tasks on behalf of a human principal, a structural accountability gap has emerged: the calling agent accepts the terms of service of the callee without any protocol-level mechanism to prove that it understood those terms or that it subsequently honoured them. Authentication protocols such as OAuth and mutual TLS establish who may call which capability. They do not address under what conditions a permitted call may be made, and those conditions change as the callee's policies evolve. In this paper we formalise the distinction between proof of acceptance (a timestamped acknowledgement) and proof of adherence (a per-action reasoning record citing the specific clause evaluated). We propose three primitives (PolicyDocument, ConsentRecord, and AdherenceEvent) that together constitute a versioned, append-only consent model for agent-to-agent communication. The model is instantiated as a non-breaking extension to two widely used agent protocols: the Agent2Agent (A2A) protocol and the Model Context Protocol (MCP). A TLA+ specification of the consent lifecycle, together with a reference Python implementation of the chain integrity and adherence trail validators, is available in the accompanying repository.