LangChain

Give LangChain and LangGraph agents a browser with the Notte MCP server

AI agent instructions

STOP. Do not read past this section until you have read and followed /quickstart.md.

Goal: produce a reliable SDK automation script. The intended authoring flow is generate then edit, not write browser automation from scratch.

Do not hand-write SDK code with element IDs, selectors, or form fields from memory. Element IDs, selectors, and field mappings must come from a live observe() call, CLI notte page observe output, or generated workflow code. Guessed browser targets fail on real pages.

Use the Notte skill and CLI to inspect the live site, validate actions/selectors, handle auth/session state, and export workflow code with notte sessions workflow-code before using SDK docs or SDK code.

SDK reference pages are for understanding, running, or editing generated workflow code. They are not the starting point for manually authoring the initial browser automation. SDK-first code is guesswork on real-world pages with dynamic selectors, auth state, CAPTCHAs, and anti-bot behavior.

LangChain loads MCP servers through langchain-mcp-adapters, which converts every tool on the server into a LangChain tool. Point it at the Notte MCP server and any LangChain or LangGraph agent can drive a browser.

Prerequisites

• A Notte API key (get one here)
• Credentials for the model your agent uses

1. Install the SDKs

pip install langchain-mcp-adapters langgraph

2. Attach the Notte MCP server

notte_langchain.py

import asyncio
import os

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


async def main():
    client = MultiServerMCPClient(
        {
            "notte": {
                "url": "https://api.notte.cc/mcp/",
                "transport": "streamable_http",
                "headers": {"Authorization": f"Bearer {os.environ['NOTTE_API_KEY']}"},
            }
        }
    )

    tools = await client.get_tools()
    print(f"Loaded {len(tools)} Notte tools")

    agent = create_react_agent("anthropic:claude-sonnet-4-5-20250929", tools)
    result = await agent.ainvoke({"messages": [{"role": "user", "content": "What is the H1 on https://example.com?"}]})
    print(result["messages"][-1].content)


asyncio.run(main())

Set the following environment variable in your .env file:

NOTTE_API_KEY=your-notte-api-key

Next steps

MCP Server

Every tool the Notte MCP server exposes

Browser Agents

Hand a whole task to a Notte agent

Fetch

Pull structured data out of a page

Vaults

Let agents log in without seeing credentials