home / mcp / grasshopper mcp server

Grasshopper MCP Server

Bridges Grasshopper with Claude Desktop via MCP to enable natural language control of Grasshopper components.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "alfredatnycu-grasshopper-mcp": {
      "command": "python",
      "args": [
        "-m",
        "grasshopper_mcp.bridge"
      ]
    }
  }
}

Grasshopper MCP Bridge lets Grasshopper on Rhino speak MCP with Claude Desktop, enabling natural language control over Grasshopper components through a dedicated bridge server. It streamlines creating and connecting components, supports high-level intent understanding, and provides guidance to ensure correct connections.

How to use

Start the Grasshopper MCP Bridge and use Claude Desktop to control Grasshopper with natural language commands. First, ensure the Grasshopper MCP Component is running inside Grasshopper so Grasshopper can receive commands. Then start the Python MCP Bridge Server so Claude Desktop can communicate with Grasshopper. In Claude Desktop, connect using the MCP protocol to the local bridge server, or configure Claude Desktop to auto-start the bridge when you launch Grasshopper.

How to install

Prerequisites you need before starting are Rhino 7 or higher, Grasshopper, Python 3.8 or higher, and Claude Desktop.

Install the Grasshopper MCP Component (GH_MCP) into Grasshopper. You have two options.

Method 1: Download the pre-compiled GH_MCP.gha file and place it in your Grasshopper libraries folder.

%APPDATA%\Grasshopper\Libraries\

Method 2: Build from source. Clone the repository and build the C# project using Visual Studio.

Install the Python MCP Bridge Server. You have three options.

Option A: Install from PyPI (recommended) using pip.

pip install grasshopper-mcp

Option B: Install from GitHub.

pip install git+https://github.com/alfredatnycu/grasshopper-mcp.git

Option C: Install from source.

git clone https://github.com/alfredatnycu/grasshopper-mcp.git
cd grasshopper-mcp
pip install -e .

If you need a specific version, install with a version tag.

pip install grasshopper-mcp==0.1.0 or
pip install git+https://github.com/alfredatnycu/[email protected]

Run the bridge server from the command line.

python -m grasshopper_mcp.bridge

Connect Claude Desktop to the MCP Bridge

Establish a connection from Claude Desktop to the bridge server using MCP. You can connect manually or configure Claude Desktop to auto-start the bridge when Claude Desktop launches.

Manual connection settings in Claude Desktop:

Protocol: MCP Host: localhost Port: 8080

Auto-start configuration for Claude Desktop (JSON snippet):

"grasshopper": {
  "command": "python",
  "args": ["-m", "grasshopper_mcp.bridge"]
}

Start using Grasshopper with Claude Desktop

With the bridge running and Claude Desktop connected, you can issue natural language commands to create, connect, and manipulate Grasshopper components. For example, describe a circle's properties or request a grid of points, and Claude Desktop will translate those instructions into Grasshopper actions through the MCP Bridge.

Example commands

Create a circle with radius 5 at point (0,0,0)

Connect the circle to an extrude component with a height of 10

Create a grid of points with 5 rows and 5 columns

Apply a random rotation to all selected objects

Troubleshooting

If issues arise, verify the GH_MCP component is loaded in Grasshopper. Check that the bridge server is running and listening on port 8080. Ensure Claude Desktop is configured to connect to localhost:8080 and that the MCP protocol is selected. Review the bridge server console for error messages.

Development and structure

The bridge consists of a Grasshopper MCP Component that provides a TCP server within Grasshopper, a Python MCP Bridge Server that connects Claude Desktop to Grasshopper, and a Component Knowledge Base stored as JSON files containing component information, patterns, and intents.

Available tools

Component creation

CreateGrasshopper components and connect them through natural language instructions using the MCP bridge.

Intent recognition

High-level understanding of user intent to generate appropriate component patterns.

Component knowledge base

JSON-backed knowledge base with component parameters, rules, and patterns.

Guidance resources

Provides guidance for Claude Desktop to correctly connect components.