Grasp (Browser Automation) MCP server

Grasp is an open-source, self-hosted agentic browser that operates in a dockerized environment. It allows AI systems to control a browser with natural language commands, providing isolated browsing capabilities that can be integrated with other AI applications through MCP and A2A support.

Installation Requirements

Before installing Grasp, ensure you have Docker installed and running on your machine.

Setting Up Grasp

Docker Installation

Run these commands to pull and start the Grasp container:

# Pull the image
docker pull getgrasp/grasp

# Run the container
docker run -d \
  --name grasp-agent \
  -p 3000:3000 \
  -e ANTHROPIC_API_KEY=YOUR_ANTHROPIC_KEY \
  getgrasp/grasp

Make sure to replace YOUR_ANTHROPIC_KEY with your actual Anthropic API key, which you can obtain from the Anthropic Console.

Accessing the Console

After starting the container, access the Grasp console by opening http://localhost:3000 in your web browser.

Usage

Integration Options

Grasp can be used in two primary ways:

Using Grasp with Claude Desktop

To use Grasp as a local MCP tool server with Claude:

1. Ensure Grasp is running in Docker
2. Configure Claude Desktop to connect to Grasp
3. Claude will then be able to operate the browser based on your instructions

Connecting an Agent to Grasp

For agent-to-agent browser automation:

1. Set up your agent to communicate via the A2A protocol
2. Configure the connection between your agent and Grasp
3. Your agent can then send commands to control the browser

Features

• Isolated Environment: Runs in Docker without affecting your local system
• Human-like Browser Automation: Controls the browser just as a human would
• Credential Handling: Can use login credentials to access personalized content
• Natural Language Control: Operates via plain language instructions without requiring code

Community Resources

For additional help and discussions, you can join the Discord community.

How to add this MCP server to Cursor

There are two ways to add an MCP server to Cursor. The most common way is to add the server globally in the ~/.cursor/mcp.json file so that it is available in all of your projects.

If you only need the server in a single project, you can add it to the project instead by creating or adding it to the .cursor/mcp.json file.

Adding an MCP server to Cursor globally

To add a global MCP server go to Cursor Settings > MCP and click "Add new global MCP server".

When you click that button the ~/.cursor/mcp.json file will be opened and you can add your server like this:

{
    "mcpServers": {
        "cursor-rules-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "cursor-rules-mcp"
            ]
        }
    }
}

Adding an MCP server to a project

To add an MCP server to a project you can create a new .cursor/mcp.json file or add it to the existing one. This will look exactly the same as the global MCP server example above.

How to use the MCP server

Once the server is installed, you might need to head back to Settings > MCP and click the refresh button.

The Cursor agent will then be able to see the available tools the added MCP server has available and will call them when it needs to.

You can also explictly ask the agent to use the tool by mentioning the tool name and describing what the function does.