Glossary
A2A Protocol
Agent-to-Agent (A2A) is an open protocol, originated by Google and now under the Linux Foundation, for peer-to-peer communication between independent agents owned by different teams or vendors. Agents publish capabilities via Agent Cards and exchange work through stateful Tasks. Distinct from MCP, which governs agent-to-tool access, not agent-to-agent.
Most systems described as "multi-agent" today are actually one team's orchestrator calling tools in sequence. A2A earns its place only when agents cross a boundary you don't own.
Do I need A2A if my agents are all built by my own team?
Usually not. A2A earns its place specifically when agents cross a boundary you don't own — a different team's agent, a different company's, or a different vendor's. If one team's orchestrator is just calling its own tools in sequence, that's not the multi-agent problem A2A was designed for, and adding the protocol adds complexity without solving anything you actually have.
Where this comes up
AI Agent Failure Modes in Production
Need more than a definition?
If this term describes a problem you have, find out where you stand with a free scored diagnostic.