A basic A2A agent

Sample using Agent-to-Agent (A2A) communication protocol

Check Agent Details

• Run the server using below command.

python main.py

• Then navigate to the URL http://localhost:9999/.well-known/agent.json
• You should be able to see Agent Details

{
  "capabilities": {

  },
  "defaultInputModes": [
    "text"
  ],
  "defaultOutputModes": [
    "text"
  ],
  "description": "A simple agent that returns weather information",
  "name": "Weather agent",
  "skills": [
    {
      "description": "Share weather specific information",
      "examples": [
        "What is the weather of blr?"
      ],
      "id": "weather_agent",
      "name": "Weather Agent",
      "tags": [
        "weather info"
      ]
    }
  ],
  "url": "http://localhost:9999",
  "version": "1.0.0"
}

Run the Client

• Once the server is up and running.
• Run the below command to get the response from the server.

python client.py

• If everything goes fine, you should a response.

{
  "id": "ccaa00f9-8e23-42bb-9880-995865b13e16",
  "jsonrpc": "2.0",
  "result": {
    "kind": "message",
    "messageId": "c0566785-ddf7-40a2-a6fe-03c0601e44e6",
    "parts": [
      {
        "kind": "text",
        "text": "{\"city\": \"BLR\", \"temp\": \"10c\"}"
      }
    ],
    "role": "agent"
  }
}

Visual Explaination

flowchart TD

    %% CLIENT SIDE
    subgraph Client_Agent
        A1["A2ACardResolver<br/>(fetches public agent card)"]
        A3["Message Payload<br/>'What's the weather in BLR?'"]
        A2["A2AClient<br/>(sends structured message)"]
    end

    %% SERVER SIDE
    subgraph Server_Agent["Server Agent (localhost:9999)"]
        B1["/.well-known/agent.json<br/>(Public Agent Card)"]
        B2["Message Receiver<br/>(POST /send-message)"]
        B3["Business Logic<br/>(Weather lookup / LLM)"]
        B4["Structured Response<br/>(JSON reply)"]
    end

    %% CONNECTIONS
    A1 --> B1
    A3 --> A2
    A2 --> B2
    B2 --> B3
    B3 --> B4
    B4 --> A2