Testing Locally

Test your agents locally before deploying.

Running Your Agent

python agent.py

The server starts on http://localhost:8080 by default.

Health Checks

# Runtime health
curl http://localhost:8080/health
# {"status": "healthy", "agents": ["my-agent"]}

# Agent health
curl http://localhost:8080/agent/my-agent/health
# {"name": "my-agent", "has_invoke": true, "has_chat": true}

Testing Invoke

curl -X POST http://localhost:8080/agent/my-agent/invoke \
  -H "Content-Type: application/json" \
  -d '{"input": {"name": "World"}, "stream": false}'

Response:

{"output": "Hello, World!"}

Testing Chat

curl -X POST http://localhost:8080/agent/my-agent/chat \
  -H "Content-Type: application/json" \
  -d '{"messages": [{"role": "user", "content": "Hello!"}], "stream": false}'

Response:

{"message": {"role": "assistant", "content": "You said: Hello!"}}

Testing Streaming

Add "stream": true to the request:

curl -X POST http://localhost:8080/agent/my-agent/invoke \
  -H "Content-Type: application/json" \
  -d '{"input": {"text": "Hello world"}, "stream": true}'

Custom Port

serve(agent, port=3000)

Or use an environment variable:

import os
serve(agent, port=int(os.environ.get("PORT", 8080)))

Multiple Agents

Test multiple agents at once:

serve([agent1, agent2], port=8080)

Each agent is available at /agent/{name}/.

Common Issues

Issue	Solution
Port already in use	Change port or kill the existing process
Connection refused	Make sure the server is running
404 Not Found	Check the agent name in the URL
501 Not Implemented	The handler (invoke/chat) isn't registered

Next Steps

Client Setup — Call deployed agents
Examples — See more examples

On this page