LangChain

The MCP adapters, turning the six tools into LangChain tools.

Setup

1. pip install langchain-mcp-adapters langgraph
2. Construct a MultiServerMCPClient pointing at the endpoint.
3. Await get_tools() and hand them to an agent.

import os, asyncio
from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent

async def main():
    client = MultiServerMCPClient({
        "unzoi": {
            "transport": "streamable_http",
            "url": "https://api.unzoi.com/mcp",
            "headers": {"x-api-key": os.environ["UNZOI_KEY"]},
        }
    })
    tools = await client.get_tools()
    agent = create_react_agent("anthropic:claude-sonnet-5", tools)
    result = await agent.ainvoke({
        "messages": [{"role": "user", "content": "What happened with port congestion this week?"}]
    })
    print(result["messages"][-1].content)

asyncio.run(main())

What differs here

• get_tools() is one round trip and the result is static. Call it at startup, not per request.
• The tools arrive with their MCP descriptions intact, which is most of why the model picks between them sensibly — do not overwrite them with shorter ones.
• A ReAct agent will happily loop. Cap its steps, or it will spend a monthly quota exploring.

Next

• Get a key at console.unzoi.com, if you have not.
• The tool reference — what the model is reading when it chooses between the six tools.
• Stories for context windows — the one habit that most changes how well an agent works against this API.
• Limits — an agent that loops can spend a monthly quota faster than you expect.
• LangChain's own MCP documentation — authoritative for anything about the client itself, including config paths, which move.

Endpoint, for copying: https://api.unzoi.com/mcp