RhinoPilot MCP

A Model Context Protocol (MCP) server for Rhino 3D, using the official @modelcontextprotocol/sdk. This server enables AI assistants and other MCP-compatible clients to interact with Rhino 3D software.

Overview

RhinoPilot MCP implements an MCP server that connects to Rhino via a secure socket connection. It provides tools for creating and manipulating 3D geometry, managing layers, capturing viewport images, and more.

Features

• MCP Server: Fully compliant with the Model Context Protocol standard using the official SDK
• Stdio Transport: Uses stdio for communication, making it compatible with various MCP clients
• Rhino Integration: Connects to the Rhino plugin via encrypted socket communication
• Geometry Creation: Create points, lines, circles, curves, NURBS surfaces, and solids
• Layer Management: Create, delete, and manage layers
• Object Selection: Select, delete, and retrieve information about objects
• Viewport Management: Capture images, list viewports, and control display modes

Prerequisites

• Node.js 18.0.0 or higher
• A running Rhino instance with the RhinoPilotHelper plugin installed

Installation

Using npx (Recommended)

You can run the RhinoPilot MCP server directly using npx without installing it:

npx @mrbeandev/rhinopilot-mcp

Global Installation

To install the package globally:

npm install -g @mrbeandev/rhinopilot-mcp

Then run it with:

rhinopilot-mcp

Local Development

1. Clone this repository

2. Install dependencies:

   npm install

3. Build the project:

   npm run build

4. Run the server:

   npm run start

Command Line Usage

Options

npx @mrbeandev/rhinopilot-mcp [options]

Available options:

• --host=<hostname>: Specify Rhino Plugin host (default: localhost)
• --port=<port>: Specify Rhino Plugin port (default: 8081)
• -h, --help: Show help message
• -v, --version: Show version information

Available Tools

The RhinoPilot MCP server provides a wide range of tools for interacting with Rhino:

Document and Connection Management

• get_document_info: Retrieve information about the current Rhino document
• check_connection: Check if the connection to the Rhino plugin is active
• connect_to_rhino: Connect to the Rhino plugin
• disconnect_from_rhino: Disconnect from the Rhino plugin

Geometry Creation

• create_point: Create a point at the specified coordinates
• create_line: Create a line between two points
• create_circle: Create a circle with specified center point and radius
• create_arc: Create an arc with center point, radius, start angle and end angle
• create_ellipse: Create an ellipse with center, x-axis endpoint and y-axis endpoint
• create_polyline: Create a polyline from a list of points
• create_curve: Create a curve from control points
• create_sphere: Create a sphere with center point and radius
• create_box: Create a box with origin, width, height, and depth
• create_cone: Create a cone with base point, height, and radius
• create_cylinder: Create a cylinder with base point, height, and radius
• create_nurbs_surface: Create a NURBS surface from a grid of control points
• create_rectangular_surface: Create a rectangular surface with origin point, width and height

Layer Management

• create_layer: Create a new layer in the document
• delete_layer: Delete a layer from the document
• get_current_layer: Get the current active layer
• get_all_layers: Get a list of all layers in the document
• set_current_layer: Set the current active layer

Object Management

• get_object_info: Get information about a specific object
• delete_object: Delete an object from the document
• select_objects: Select objects in the document
• get_selected_objects_info: Get information about the currently selected objects

Visual and Display Management

• capture_viewport_image: Capture an image of the current viewport
• get_viewport_list: Get a list of available viewports
• get_display_modes: Get a list of available display modes
• set_viewport_display_mode: Set the display mode for a specific viewport

Scripting

• execute_rhino_script: Execute a Rhino script command

Example Usage

Here's how to use the tools with an MCP client:

// Using an MCP client to connect to the server
const result = await client.callTool({
  name: "create_box",
  arguments: {
    origin: [0, 0, 0],
    width: 10,
    height: 5,
    depth: 3
  }
});

// Create a layer and set it as current
await client.callTool({
  name: "create_layer",
  arguments: {
    name: "MyNewLayer",
    color: [255, 0, 0]  // Red color
  }
});

await client.callTool({
  name: "set_current_layer",
  arguments: {
    name: "MyNewLayer"
  }
});

License

This project is licensed under the MIT License.