LangGraph Integration with MCP Gateway¶

LangGraph is a framework for developing applications powered by language models. Integrating LangGraph with the Model Context Protocol (MCP) allows agents to utilize tools defined across one or more MCP servers, enabling seamless interaction with external data sources and services.

🧰 Key Features¶

Dynamic Tool Access: Connects to MCP servers to fetch available tools in real time.
Multi-Server Support: Interact with tools defined on multiple MCP servers simultaneously.
Standardized Communication: Utilizes the open MCP standard for consistent tool integration.

🛠 Installation¶

To use MCP tools in LangGraph, install the langchain-mcp-adapters package:

pip install langchain-mcp-adapters

🔗 Connecting to MCP Gateway¶

Here's how to set up a connection to your MCP Gateway:

from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent

client = MultiServerMCPClient(
    {
        "gateway": {
            "url": "http://localhost:4444/mcp",
            "transport": "streamable_http",
        }
    }
)

Replace "http://localhost:4444/mcp" with the URL of your MCP Gateway.

🤖 Creating an Agent¶

After setting up the client, you can create a LangGraph agent:

agent = create_react_agent(
    tools=client.get_tools(),
    llm=your_language_model,
)

Replace your_language_model with your configured language model instance.

🧪 Using the Agent¶

Once the agent is created, you can use it to perform tasks:

response = agent.run("Use the 'weather' tool to get the forecast for Dublin.")
print(response)