LangChain / LangGraph · CodeIntegrity Orchestrator

Use langchain-mcp-adapters to load your gateway's tools into a LangChain or LangGraph agent over Streamable HTTP.

Copy the gateway URL, create a Gateway API Key in MCP Gateways, and export both values:

export CODEINTEGRITY_MCP_URL='https://your-host/mcp/{slug}'
export CODEINTEGRITY_API_KEY='ci_mcp_...'

pip install langchain-mcp-adapters "langchain[openai]"

import os

from langchain.agents import create_agent
from langchain_mcp_adapters.client import MultiServerMCPClient

client = MultiServerMCPClient(
    {
        "orchestrator": {
            "transport": "http",
            "url": os.environ["CODEINTEGRITY_MCP_URL"],
            "headers": {
                "Authorization": f"Bearer {os.environ['CODEINTEGRITY_API_KEY']}"
            },
        }
    }
)

# inside an async context
tools = await client.get_tools()
agent = create_agent("openai:gpt-4.1", tools)
result = await agent.ainvoke({"messages": "What tools can you use?"})

The gateway exposes search, execute_code, and authorize. Programmatic agents normally authenticate with the Gateway API Key above.