Glossary

Agent-to-agent (A2A)

Agent-to-agent (A2A) is an open protocol that lets AI agents built by different teams find each other and pass work between them. Each agent describes itself in an agent card: what it can do and how to call it.

In a run

When one agent hands work to another over A2A, the call carries a credential that the receiving side checks. Content is inspected in both directions, the permission narrows at the handoff, and the record shows both agents.

Our word for it

Governed communication: a controlled path between agents.

What Ares adds to the protocol

A2A carries the message between agents. Ares decides what the receiving agent may do with it. Every agent-to-agent call gets the same access policy check and the same record as a call to a tool. The record holds both sides of the handoff: the calling agent, the receiving agent, and the permission after narrowing. This holds across companies: when work crosses to a supplier’s agent, the rules follow it. See across organisations.

Bring the agents you have

Your existing agents connect over A2A, MCP or our SDK, without a rewrite. A2A is for agent to agent; MCP is for agent to tool.

Check your agent card

The free agent card validator checks an A2A card and maps the permissions it asks for. Run it before you enrol the agent, so you know up front what the card asks for.

Where it stops

Ares covers the agents you enrol. Two agents that talk over A2A outside Ares are not checked and are not on the record.

The two-company sandbox run shows a handoff narrow at the company boundary.

Related terms

  • Model Context Protocol (MCP)The Model Context Protocol (MCP) is an open protocol for connecting AI agents to tools and data sources.
  • DelegationDelegation is one party handing a task, and part of its authority, to another.

Tell us about one workflow.

We will say whether an agent should do it, and what the first step costs.

A senior member of the team replies within two business days.