What is the A2A (Agent-to-Agent) protocol in Google ADK? How does it enable inter-agent communication?

A2A (Agent-to-Agent) Protocol

The Agent-to-Agent (A2A) protocol is an open standard developed by Google that enables AI agents built with different frameworks to communicate and collaborate with each other over the network.

Why A2A?

How A2A Works

Key Components

1. Agent Card

A JSON document that advertises an agent's capabilities:

jsonCopy
{
  "name": "research-agent",
  "description": "Researches topics using web search",
  "url": "https://research-agent.example.com",
  "skills": [
    {
      "id": "web-search",
      "name": "Web Search",
      "description": "Search the web for information"
    }
  ],
  "authentication": {
    "schemes": ["bearer"]
  }
}

2. Task Lifecycle

A2A in Google ADK

pythonCopy
from google.adk.agents import Agent
from google.adk.a2a import A2AClient, A2AServer

# Server: Expose an agent via A2A
server_agent = Agent(
    name="research_agent",
    model="gemini-2.5-flash",
    instruction="You research topics using web search.",
    tools=[google_search],
)

# Serve over A2A protocol
server = A2AServer(agent=server_agent, port=8080)
server.start()

# Client: Discover and call the remote agent
client = A2AClient(url="https://research-agent.example.com")
card = client.discover()
result = client.send_task("What is the latest news on LLaMA 4?")

A2A vs MCP

Key insight: A2A and MCP are complementary. MCP connects agents to tools, A2A connects agents to agents.

Learn more at A2A Protocol and ADK A2A Docs.