OpenAgenet/OAN: A Trust Layer for Secure Multi-Agent Networks

The paper introduces OpenAgenet (OAN), an open infrastructure for trusted agent interconnection, addressing a critical gap as agents move from isolated applications into open, multi-operator networks. Before an agent can safely discover, select, and invoke another agent, it needs to verify identity provenance, governance state, discovery authorization, freshness, and pre-connection trust evidence. OAN is designed as a protocol-neutral trust layer that does not replace existing agent interaction protocols (like MCP, A2A, or ANP) but layers on top of them. The architecture includes Root-governed identity admission, Registrar-assisted onboarding, Root-verified package publication, authorization-aware Discovery, and signed trusted invocation.

OAN introduces a governance model with roles for Root authority, Registrars, and Agents, and uses a blockchain-backed authorization bulletin for tamper-evident trust records. The paper covers deployment patterns, cooperation models, prototype status, and performance benchmarks. The project is published on arXiv (arXiv:2606.03161) and aims to enable secure, scalable agent ecosystems across different operators. This is a foundational piece for the emerging multi-agent internet, where trust and interoperability are paramount.