How to integrate Browser tool MCP with Autogen

BOOK A CALLGET STARTED FOR FREE
Framework Integration Gradient
Browser tool Logo
AutoGen Logo
divider

Introduction

This guide walks you through connecting Browser tool to AutoGen using the Composio tool router. By the end, you'll have a working Browser tool agent that can copy highlighted text from this webpage, drag and drop a file to upload section, fetch and summarize main page content, click login button at top right corner through natural language commands.

This guide will help you understand how to give your AutoGen agent real control over a Browser tool account through Composio's Browser tool MCP server.

Before we dive in, let's take a quick look at the key ideas and tools involved.

TL;DR

Here's what you'll learn:
  • Get and set up your OpenAI and Composio API keys
  • Install the required dependencies for Autogen and Composio
  • Initialize Composio and create a Tool Router session for Browser tool
  • Wire that MCP URL into Autogen using McpWorkbench and StreamableHttpServerParams
  • Configure an Autogen AssistantAgent that can call Browser tool tools
  • Run a live chat loop where you ask the agent to perform Browser tool operations

What is AutoGen?

Autogen is a framework for building multi-agent conversational AI systems from Microsoft. It enables you to create agents that can collaborate, use tools, and maintain complex workflows.

Key features include:

  • Multi-Agent Systems: Build collaborative agent workflows
  • MCP Workbench: Native support for Model Context Protocol tools
  • Streaming HTTP: Connect to external services through streamable HTTP
  • AssistantAgent: Pre-built agent class for tool-using assistants

What is the Browser tool MCP server, and what's possible with it?

The Browser tool MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to browser automation tools. It provides structured and secure access to browser actions, so your agent can fetch web content, perform clicks, automate keyboard shortcuts, move the mouse, and interact with on-page elements just like a real user.

  • Fetch and analyze webpage content: Let your agent retrieve the full HTML or clean text of any web page for data extraction, analysis, or decision-making.
  • Automated mouse and keyboard interactions: Instruct your agent to perform precise clicks, double clicks, drags, and keyboard shortcuts to navigate, select, or manipulate content on the page.
  • Clipboard and text extraction: Have the agent copy highlighted text, read clipboard contents, or transfer data between the browser and other tools for seamless workflows.
  • Drag-and-drop automation: Enable your agent to handle complex drag-and-drop actions, such as moving files or rearranging lists, to mimic advanced user interactions.
  • Fine-grained UI element control: Direct your agent to move the mouse, press and hold, or release buttons at exact coordinates to interact with dynamic or custom web interfaces.

Supported Tools & Triggers

Tools
Copy Selected TextCopy currently selected text on the page to clipboard - ideal for extracting highlighted content, copying form data, or harvesting visible text selections.
Drag and DropExecute precise drag and drop operations - essential for file uploads, list reordering, element moving, and complex ui interactions that require drag-based manipulation.
Fetch Webpage ContentYour eyes: get page content for decision-making.
Get Clipboard ContentRead current content from the system clipboard - essential for data transfer workflows, extracting copied text, and reading user-copied data for processing.
Keyboard ShortcutExecute keyboard shortcuts and key combinations - essential for copy/paste, navigation, and application commands that agents need for efficient browser automation.
Mouse ClickPrecision clicker: manual clicking with coordinates.
Mouse Double ClickExecute a precise double click at specified screen coordinates - ideal for opening files, selecting text, or activating ui elements that require double click gestures.
Mouse Down (Press and Hold)Press and hold mouse button at coordinates - use for starting custom drag operations, text selections, or long-press interactions.
Mouse MoveMove mouse cursor to precise coordinates without clicking - perfect for triggering hover effects, revealing tooltips, and positioning for subsequent interactions.
Mouse Up (Release Button)Release mouse button at coordinates - completes drag operations, text selections, and long-press interactions.
Navigate to URLAlways start here: creates browser session and navigates to url.
Paste TextPaste text content at the current cursor position - perfect for filling forms, inserting data into text fields, or quick content insertion at focused elements.
AI Perform Web TaskAi automation: complex workflows only.
Screenshot WebpageCapture high-quality screenshot of any webpage with extensive customization options - perfect for archiving, visual documentation, full-page captures, and cross-device viewport testing.
Scroll PagePage navigation: smooth scrolling.
Set Clipboard ContentStore text content in the system clipboard for later paste operations - perfect for preparing data transfers, staging content for forms, or cross-application data sharing.
Take ScreenshotVisual verification: capture screenshot of current browser viewport.
Type TextControlled input: human-like typing.

What is the Composio tool router, and how does it fit here?

What is Tool Router?

Composio's Tool Router helps agents find the right tools for a task at runtime. You can plug in multiple toolkits (like Gmail, HubSpot, and GitHub), and the agent will identify the relevant app and action to complete multi-step workflows. This can reduce token usage and improve the reliability of tool calls. Read more here: Getting started with Tool Router

The tool router generates a secure MCP URL that your agents can access to perform actions.

How the Tool Router works

The Tool Router follows a three-phase workflow:

  1. Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
  2. Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
  3. Execution: Executes the action using the authenticated connection.

Step-by-step Guide

Prerequisites

You will need:

  • A Composio API key
  • An OpenAI API key (used by Autogen's OpenAIChatCompletionClient)
  • A Browser tool account you can connect to Composio
  • Some basic familiarity with Autogen and Python async

Getting API Keys for OpenAI and Composio

OpenAI API Key
  • Go to the OpenAI dashboard and create an API key. You'll need credits to use the models, or you can connect to another model provider.
  • Keep the API key safe.
Composio API Key
  • Log in to the Composio dashboard.
  • Navigate to your API settings and generate a new API key.
  • Store this key securely as you'll need it for authentication.

Install dependencies

bash
pip install composio python-dotenv
pip install autogen-agentchat autogen-ext-openai autogen-ext-tools

Install Composio, Autogen extensions, and dotenv.

What's happening:

  • composio connects your agent to Browser tool via MCP
  • autogen-agentchat provides the AssistantAgent class
  • autogen-ext-openai provides the OpenAI model client
  • autogen-ext-tools provides MCP workbench support

Set up environment variables

bash
COMPOSIO_API_KEY=your-composio-api-key
OPENAI_API_KEY=your-openai-api-key
USER_ID=your-user-identifier@example.com

Create a .env file in your project folder.

What's happening:

  • COMPOSIO_API_KEY is required to talk to Composio
  • OPENAI_API_KEY is used by Autogen's OpenAI client
  • USER_ID is how Composio identifies which user's Browser tool connections to use

Import dependencies and create Tool Router session

python
import asyncio
import os
from dotenv import load_dotenv
from composio import Composio

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_ext.tools.mcp import McpWorkbench, StreamableHttpServerParams

load_dotenv()

async def main():
    # Initialize Composio and create a Browser tool session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["browser_tool"]
    )
    url = session.mcp.url
What's happening:
  • load_dotenv() reads your .env file
  • Composio(api_key=...) initializes the SDK
  • create(...) creates a Tool Router session that exposes Browser tool tools
  • session.mcp.url is the MCP endpoint that Autogen will connect to

Configure MCP parameters for Autogen

python
# Configure MCP server parameters for Streamable HTTP
server_params = StreamableHttpServerParams(
    url=url,
    timeout=30.0,
    sse_read_timeout=300.0,
    terminate_on_close=True,
    headers={"x-api-key": os.getenv("COMPOSIO_API_KEY")}
)

Autogen expects parameters describing how to talk to the MCP server. That is what StreamableHttpServerParams is for.

What's happening:

  • url points to the Tool Router MCP endpoint from Composio
  • timeout is the HTTP timeout for requests
  • sse_read_timeout controls how long to wait when streaming responses
  • terminate_on_close=True cleans up the MCP server process when the workbench is closed

Create the model client and agent

python
# Create model client
model_client = OpenAIChatCompletionClient(
    model="gpt-5",
    api_key=os.getenv("OPENAI_API_KEY")
)

# Use McpWorkbench as context manager
async with McpWorkbench(server_params) as workbench:
    # Create Browser tool assistant agent with MCP tools
    agent = AssistantAgent(
        name="browser_tool_assistant",
        description="An AI assistant that helps with Browser tool operations.",
        model_client=model_client,
        workbench=workbench,
        model_client_stream=True,
        max_tool_iterations=10
    )

What's happening:

  • OpenAIChatCompletionClient wraps the OpenAI model for Autogen
  • McpWorkbench connects the agent to the MCP tools
  • AssistantAgent is configured with the Browser tool tools from the workbench

Run the interactive chat loop

python
print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
print("Ask any Browser tool related question or task to the agent.\n")

# Conversation loop
while True:
    user_input = input("You: ").strip()

    if user_input.lower() in ["exit", "quit", "bye"]:
        print("\nGoodbye!")
        break

    if not user_input:
        continue

    print("\nAgent is thinking...\n")

    # Run the agent with streaming
    try:
        response_text = ""
        async for message in agent.run_stream(task=user_input):
            if hasattr(message, "content") and message.content:
                response_text = message.content

        # Print the final response
        if response_text:
            print(f"Agent: {response_text}\n")
        else:
            print("Agent: I encountered an issue processing your request.\n")

    except Exception as e:
        print(f"Agent: Sorry, I encountered an error: {str(e)}\n")
What's happening:
  • The script prompts you in a loop with You:
  • Autogen passes your input to the model, which decides which Browser tool tools to call via MCP
  • agent.run_stream(...) yields streaming messages as the agent thinks and calls tools
  • Typing exit, quit, or bye ends the loop

Complete Code

Here's the complete code to get you started with Browser tool and AutoGen:

import asyncio
import os
from dotenv import load_dotenv
from composio import Composio

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_ext.tools.mcp import McpWorkbench, StreamableHttpServerParams

load_dotenv()

async def main():
    # Initialize Composio and create a Browser tool session
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    session = composio.create(
        user_id=os.getenv("USER_ID"),
        toolkits=["browser_tool"]
    )
    url = session.mcp.url

    # Configure MCP server parameters for Streamable HTTP
    server_params = StreamableHttpServerParams(
        url=url,
        timeout=30.0,
        sse_read_timeout=300.0,
        terminate_on_close=True,
        headers={"x-api-key": os.getenv("COMPOSIO_API_KEY")}
    )

    # Create model client
    model_client = OpenAIChatCompletionClient(
        model="gpt-5",
        api_key=os.getenv("OPENAI_API_KEY")
    )

    # Use McpWorkbench as context manager
    async with McpWorkbench(server_params) as workbench:
        # Create Browser tool assistant agent with MCP tools
        agent = AssistantAgent(
            name="browser_tool_assistant",
            description="An AI assistant that helps with Browser tool operations.",
            model_client=model_client,
            workbench=workbench,
            model_client_stream=True,
            max_tool_iterations=10
        )

        print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
        print("Ask any Browser tool related question or task to the agent.\n")

        # Conversation loop
        while True:
            user_input = input("You: ").strip()

            if user_input.lower() in ['exit', 'quit', 'bye']:
                print("\nGoodbye!")
                break

            if not user_input:
                continue

            print("\nAgent is thinking...\n")

            # Run the agent with streaming
            try:
                response_text = ""
                async for message in agent.run_stream(task=user_input):
                    if hasattr(message, 'content') and message.content:
                        response_text = message.content

                # Print the final response
                if response_text:
                    print(f"Agent: {response_text}\n")
                else:
                    print("Agent: I encountered an issue processing your request.\n")

            except Exception as e:
                print(f"Agent: Sorry, I encountered an error: {str(e)}\n")

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

Conclusion

You now have an Autogen assistant wired into Browser tool through Composio's Tool Router and MCP. From here you can:
  • Add more toolkits to the toolkits list, for example notion or hubspot
  • Refine the agent description to point it at specific workflows
  • Wrap this script behind a UI, Slack bot, or internal tool
Once the pattern is clear for Browser tool, you can reuse the same structure for other MCP-enabled apps with minimal code changes.

How to build Browser tool MCP Agent with another framework

OpenAI Agents SDK

OpenAI Agents SDK

Use Browser tool MCP with OpenAI Agents SDK

Claude Agent SDK

Claude Agent SDK

Use Browser tool MCP with Claude Agent SDK

Google ADK

Google ADK

Use Browser tool MCP with Google ADK

LangChain

LangChain

Use Browser tool MCP with LangChain

Vercel AI SDK

Vercel AI SDK

Use Browser tool MCP with Vercel AI SDK

Mastra AI

Mastra AI

Use Browser tool MCP with Mastra AI

LlamaIndex

LlamaIndex

Use Browser tool MCP with LlamaIndex

CrewAI

CrewAI

Use Browser tool MCP with CrewAI

FAQ

What are the differences in Tool Router MCP and Browser tool MCP?

With a standalone Browser tool MCP server, the agents and LLMs can only access a fixed set of Browser tool tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Browser tool and many other apps based on the task at hand, all through a single MCP endpoint.

Can I use Tool Router MCP with Autogen?

Yes, you can. Autogen fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Browser tool tools.

Can I manage the permissions and scopes for Browser tool while using Tool Router?

Yes, absolutely. You can configure which Browser tool scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do.

How safe is my data with Composio Tool Router?

All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Browser tool data and credentials are handled as safely as possible.

Used by agents from

Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai

Never worry about agent reliability

We handle tool reliability, observability, and security so you never have to second-guess an agent action.

Book a call↗Get started for free