Agents
A2A protocol (Agent2Agent)
Open protocol introduced by Google in 2025 for interoperability and communication between AI agents built on different frameworks and by different vendors.
A2A protocol (Agent2Agent) is an open standard, announced by Google in April 2025, for communication and interoperability between AI agents built on different frameworks and by different vendors. It lets agents discover one another, delegate tasks, and exchange results even when they share no memory, tools, or internal context.
Technically, each agent publishes an Agent Card — a JSON file (typically at /.well-known/agent-card.json) describing its name, capabilities, service endpoint, and authentication requirements. Communication runs over existing web standards: HTTP, SSE, and JSON-RPC, with security based on TLS, JWTs, and OpenID Connect. The protocol supports long-running tasks and is modality-agnostic (text, audio, video). Agents remain “opaque,” collaborating without exposing their internal logic.
A2A is positioned as a complement to MCP: where MCP connects agents to tools and data sources, A2A connects agents to each other. In June 2025 Google donated the protocol to the Linux Foundation for vendor-neutral governance, with backing from over a hundred companies, making it a leading candidate standard for multi-agent systems.