Skip to content

Getting Started

Installation

pip install a2akit

pip install a2akit[langgraph]

pip install a2akit[postgres]

pip install a2akit[sqlite]

Your First Agent

Create a file called my_agent.py:

from a2akit import A2AServer, AgentCardConfig, TaskContext, Worker


class EchoWorker(Worker):  # (1)!
    async def handle(self, ctx: TaskContext) -> None:  # (2)!
        await ctx.complete(f"Echo: {ctx.user_text}")  # (3)!


server = A2AServer(
    worker=EchoWorker(),
    agent_card=AgentCardConfig(  # (4)!
        name="Echo Agent",
        description="Echoes your input back.",
        version="0.1.0",
    ),
)
app = server.as_fastapi_app()  # (5)!
  1. Subclass Worker to define your agent's behavior.
  2. Implement the single required method: handle(ctx).
  3. Call a lifecycle method to complete the task with a text artifact.
  4. Configure the agent discovery card (served at /.well-known/agent-card.json).
  5. Get a fully configured FastAPI app — storage, broker, event bus, and all endpoints wired automatically.

Start the Server

uvicorn my_agent:app --reload

Your agent is now running at http://localhost:8000 with all A2A endpoints active.

Test It

curl -X POST http://localhost:8000/v1/message:send \
  -H "Content-Type: application/json" \
  -d '{"message":{"role":"user","parts":[{"text":"hello"}],"messageId":"1"}}'

Expected response:

{
  "id": "...",
  "contextId": "...",
  "status": {
    "state": "completed",
    "timestamp": "..."
  },
  "artifacts": [
    {
      "artifactId": "final-answer",
      "parts": [{"text": "Echo: hello"}]
    }
  ],
  "history": [...]
}

Discover the Agent

curl http://localhost:8000/.well-known/agent-card.json | python -m json.tool

This returns the agent's discovery card with capabilities, skills, and supported transport protocols.

What's Next?