Agent-to-agent communication lets one agent delegate, coordinate, request information, or pass results to another. Messages can be structured protocol objects, tool calls, shared-state updates, or ordinary text.
Every message should be treated as untrusted input unless its authority and scope are verified. Receiving agents need to distinguish information from instructions and reject requests that conflict with their assigned task or permission boundary.
ELI5
Agent-to-agent communication lets AI agents exchange information directly using an agreed format. This can save a person from copying the same details between separate services.
For example, a shopping agent can send a merchant agent the required size, budget, and delivery date. The merchant agent can reply with matching stock, while the shopping agent still checks the facts before recommending anything.
