Skip to main content
This example demonstrates basic remote MCP (Model Context Protocol) server usage with the Dedalus SDK for connecting to external tools and services. 
import asyncio
from dedalus_labs import AsyncDedalus, DedalusRunner
from dotenv import load_dotenv
from dedalus_labs.utils.stream import stream_async

load_dotenv()

async def main():
    client = AsyncDedalus()
    runner = DedalusRunner(client)

    result = await runner.run(
        input="Who won Wimbledon 2025?",
        model="openai/gpt-5-mini",
        mcp_servers=["simon-liang/brave-search-mcp"],
        stream=False
    )

    print(result.final_output)

if __name__ == "__main__":
    asyncio.run(main())

​
Using Local MCP Servers

If you want to use a locally running MCP server, you’ll need to expose it via a public tunnel since Dedalus needs to access your server over the internet.
Install Cloudflare Tunnel: 
brew install cloudflared
Start the tunnel: 
# Replace the port with wherever your MCP server is running
cloudflared tunnel --url http://localhost:8000
This will output a public URL (e.g., https://random-name.trycloudflare.com) that you can use to connect your Dedalus agent to your local MCP server. 
import asyncio
from dedalus_labs import AsyncDedalus, DedalusRunner
from dotenv import load_dotenv
from dedalus_labs.utils.stream import stream_async

load_dotenv()

async def main():
    client = AsyncDedalus()
    runner = DedalusRunner(client)

    result = await runner.run(
        input="Who won Wimbledon 2025?",
        model="openai/gpt-5-mini",
        mcp_servers="https://random-name.trycloudflare.com",
        stream=False
    )

    print(result.final_output)

if __name__ == "__main__":
    asyncio.run(main())