Skip to content

Getting Started

Welcome to the Napari MCP Server! This section will help you get up and running with AI-controlled napari in just a few minutes.

Choose Your Path

Two simple ways to set up napari MCP:

Method Description Best For Setup Time
⚡ Quick Start CLI installer automatically configures your AI app Most users ~3 minutes
🛠️ Advanced Installation Manual configuration and development setup Advanced users, developers ~10 minutes

What You'll Need

  • Python 3.10+
  • pip (comes with Python)

For Advanced Installation

  • Python 3.10+
  • pip or uv
  • Git (for development install)

Installation Overview

# 1. Install package
pip install napari-mcp

# 2. Auto-configure your application
napari-mcp-install claude-desktop  # or claude-code, cursor, etc.

# 3. Restart application and start using!

Advanced Method

For manual configuration, zero-install with uv, or development setup: - Zero Install - Use uv without permanent installation - Manual Configuration - Detailed setup for all platforms - Development Setup - Clone repository and install in editable mode

Expected Timeline

Method Setup Time Use Case
Quick Start (CLI) ~3 minutes First-time users, quick demo
Manual Config ~5 minutes Custom configurations
Zero Install ~2 minutes Testing, CI/CD pipelines
Development ~10 minutes Contributing, plugin development

Success Indicators

After following any guide, you should see:

  • No errors during installation/configuration
  • Napari window opens when AI app requests it
  • Your AI app can call session_information() successfully
  • Screenshot tool works and returns images

Supported AI Applications

The CLI installer supports:

Application Command Platform
Claude Desktop napari-mcp-install claude-desktop macOS, Windows, Linux
Claude Code napari-mcp-install claude-code macOS, Windows, Linux
Cursor IDE napari-mcp-install cursor macOS, Windows, Linux
Cline (VS Code) napari-mcp-install cline-vscode macOS, Windows, Linux
Cline (Cursor) napari-mcp-install cline-cursor macOS, Windows, Linux
Gemini CLI napari-mcp-install gemini macOS, Windows, Linux
Codex CLI napari-mcp-install codex macOS, Windows, Linux
All napari-mcp-install all Install for all apps

→ See Integration Guides for app-specific details

Need Help?

If you run into issues:

  1. Check prerequisites - Ensure Python 3.10+ and pip are installed
  2. Verify installation - Run napari-mcp-install --help
  3. Restart AI apps - Claude Desktop, Cursor, etc. after config changes
  4. Check logs - Look for error messages in terminal output

→ See our Troubleshooting Guide for detailed help


Ready to begin? Pick your preferred method above and let's get started! 🚀