MCP-Slicer - 3D Slicer Model Context Protocol Integration

English | 简体中文

MCP-Slicer connects 3D Slicer with model clients like Claude Desktop or Cline through the Model Context Protocol (MCP), enabling direct interaction and control of 3D Slicer. This integration allows for medical image processing, scene creation, and manipulation using natural language.

Features

1. list_nodes: List and filter Slicer MRML nodes and view their properties

2. execute_python_code: Execute Python code in the Slicer environment

3. capture_screenshot: Capture real-time screenshots of Slicer views

Installation

Prerequisites

• 3D Slicer 5.8 or newer
• Python 3.13 or newer
• uv package manager

If you're on Mac, please install uv as

brew install uv

On Windows

powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

and then

set Path=C:\Users\nntra\.local\bin;%Path%

Otherwise installation instructions are on their website: Install uv

⚠️ Please install UV first

Claude for Desktop Integration

Go to Claude > Settings > Developer > Edit Config > claude_desktop_config.json to include the following:

{
  "mcpServers": {
    "slicer": {
      "command": "uvx",
      "args": ["mcp-slicer"]
    }
  }
}

Cline Intergration

{
  "mcpServers": {
    "slicer": {
      "command": "uvx",
      "args": ["mcp-slicer"]
    }
  }
}

Usage

Check Claude Settings

Make sure you see the corresponding slicer tools added to the Claude Desktop App

Open Slicer Web Server

1. Open the Slicer Web Server module,

2. ensure the required interfaces are checked,

3. then start the server

Examples

- list_nodes

> What Markups nodes are in the Slicer scene now, list their names, what is their length if it is a line, and what is its angle if it is an angle

- execute python code

> Draw a translucent green cube of 8 cm in the Slicer scene, mark its vertices, and then draw a red sphere inscribed in it.

- capture_screenshot

> Capture the current state of Slicer to provide visual feedback to AI

Usage examples:

• `capture_screenshot()` - Capture full application window
• `capture_screenshot(view_type="slice", view_name="red")` - Capture Red slice view
• `capture_screenshot(view_type="3d", camera_axis="A")` - Capture 3D view from anterior

This enables a complete REACT loop where AI can:

1. Reason about what to do

2. Act using `execute_python_code`

3. Observe the result using `capture_screenshot`

Technical Details

Utilizes existing Slicer Web Server interfaces. For technical details, please see Slicer web server user guide

Limitations & Security Considerations

• The `execute_python_code` tool allows running arbitrary Python code in 3D Slicer, which is powerful but potentially dangerous.

⚠️ Not recommended for production use.

• Complex operations may need to be broken down into smaller steps.

Contributing

Contributions are welcome! Feel free to submit Pull Requests.

Disclaimer

This is a third-party integration project, not developed by the 3D Slicer team.