A2A (Agent-to-Agent Protocol)

A2A (Agent-to-Agent Protocol) is an open protocol developed by Google that enables AI agents built on different frameworks and by different vendors to communicate with each other seamlessly. A2A complements Anthropic's Model Context Protocol (MCP) by focusing on agent-to-agent interoperability rather than tool access.

A2A allows agents to discover each other's capabilities through standardized Agent Cards, negotiate communication modalities (text, forms, files, streams), and collaborate on complex tasks. The protocol supports enterprise-grade security, including authentication via OAuth and enterprise identity providers.

A2A is one of several complementary protocols forming the infrastructure stack of the Internet of Agents. MIT's NANDA Protocol bridges A2A, MCP, and Microsoft's NLWeb through a protocol adapter, providing the meta-layer for cross-protocol agent discovery, identity verification, and decentralized routing. In this architecture, A2A handles direct agent communication, MCP handles tool connections, and NANDA provides the discovery and trust fabric across all of them.

A2A represents a critical piece of the emerging agentic infrastructure, enabling multi-agent workflows where specialized agents can delegate subtasks to one another without requiring shared memory or identical frameworks.