Philips Hue MCP Server

A powerful Model Context Protocol (MCP) interface for controlling Philips Hue smart lighting systems. Enable AI assistants like Claude to control your lights using natural language.

Table of Contents

• Philips Hue MCP Server
• Table of Contents
• Overview
• Features
• Quick Start
• Setup
• Prerequisites
• Installation
• First Run
• Using with Claude
• Install in Claude Desktop
• Manual Configuration
• Test with MCP Inspector
• API Reference
• Resources
• Tools
• Prompts
• Examples
• Controlling Single Lights
• Working with Groups
• Creating Scenes
• Advanced Options
• Command Line Arguments
• Development Mode
• Troubleshooting
• How It Works
• Contributing
• Project Structure
• Development
• License

Overview

This server leverages the Model Context Protocol (MCP) to provide a seamless integration between AI assistants like Claude and your Philips Hue lighting system. With it, you can control your smart lights using natural language, access detailed lighting information, and create advanced lighting setups through a standardized AI-friendly interface.

Features

• Complete Light Control: Turn on/off, adjust brightness, change colors, set color temperature
• Comprehensive Group Management: Control multiple lights together, create custom groups
• Scene Handling: Apply existing scenes, create quick custom lighting scenes
• Activity-Based Presets: Ready-made settings for reading, relaxation, concentration, and more
• Special Effects: Access dynamic lighting effects like color loops
• Natural Language Control: Specialized prompts for lighting control through conversation
• Secure Local Integration: Connects directly to your Hue bridge on your local network

Quick Start

# Install dependencies using uv (recommended)
uv sync

# Test with MCP Inspector
uv run mcp dev hue_server.py

# Install in Claude Desktop
uv run mcp install hue_server.py --name "Philips Hue"

Then in Claude, start with: "I'd like to control my Philips Hue lights. Can you show me which lights I have available?"

Setup

Prerequisites

• Python 3.10 or higher
• uv package manager (recommended)
• A Philips Hue bridge on your local network
• Philips Hue lights paired with your bridge

Installation

Using uv (recommended):

# Clone the repository
git clone https://github.com/ThomasRohde/hue-mcp.git
cd hue-mcp

# Install dependencies and create virtual environment automatically
uv sync

# Activate the virtual environment (optional, uv run handles this automatically)
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

Using pip:

# Clone the repository
git clone https://github.com/ThomasRohde/hue-mcp.git
cd hue-mcp

# Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -e ".[dev]"

First Run

1. Test the server with the MCP Inspector:

uv run mcp dev hue_server.py

1. When prompted, press the link button on your Hue bridge to authorize the connection
2. Your connection details will be saved in ~/.hue-mcp/config.json for future use

Using with Claude

Install in Claude Desktop

The easiest way to use this server with Claude Desktop:

# Install with default name
uv run mcp install hue_server.py

# Or with custom name
uv run mcp install hue_server.py --name "Philips Hue Controller"

# With environment variables if needed
uv run mcp install hue_server.py --name "Hue" -v DEBUG=1

Manual Configuration

Alternatively, you can manually configure Claude Desktop by editing the configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the following configuration:

macOS:

{
  "mcpServers": {
    "hue": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/username/Projects/hue-mcp",
        "run",
        "hue_server.py"
      ]
    }
  }
}

Windows:

{
  "mcpServers": {
    "hue": {
      "command": "uv",
      "args": [
        "--directory",
        "c:\\Users\\username\\Projects\\hue-mcp",
        "run",
        "hue_server.py"
      ]
    }
  }
}

Replace /Users/username/Projects/hue-mcp (macOS) or c:\\Users\\username\\Projects\\hue-mcp (Windows) with the actual path to your cloned repository. The --directory flag ensures uv runs in the correct project directory and can find the dependencies defined in pyproject.toml.

After updating the configuration, restart Claude Desktop for the changes to take effect.

Test with MCP Inspector

For development and testing, use the MCP Inspector:

uv run mcp dev hue_server.py

# With additional dependencies
uv run mcp dev hue_server.py --with pandas

# With debug logging
uv run mcp dev hue_server.py --log-level debug

API Reference

Resources

Tools

Prompts

Examples

Controlling Single Lights

# Turn on a light
turn_on_light(1)

# Set a light to 50% brightness
set_brightness(1, 127)

# Change a light color to purple
set_color_rgb(1, 128, 0, 128)

# Set reading mode
set_color_preset(1, "reading")

Working with Groups

# Turn off all lights in living room (group 2)
turn_off_group(2)

# Create a new group
create_group("Bedroom", [3, 4, 5])

# Set all kitchen lights to energizing mode
set_group_color_preset(3, "energize")

Creating Scenes

# Apply an existing scene
set_scene(2, "abc123")  # Group 2, scene ID abc123

# Create a quick relaxing scene for the living room
quick_scene("Evening Relaxation", group_id=2, rgb=[255, 147, 41], brightness=120)

Advanced Options

Command Line Arguments

The server supports the following command line arguments when run directly:

# Run with stdio transport (default, for MCP clients)
python hue_server.py

# Run with custom host and port (HTTP/SSE mode)
python hue_server.py --sse --host 0.0.0.0 --port 8888

# Enable debug logging
python hue_server.py --log-level debug

# Show all available options
python hue_server.py --help

Development Mode

When developing or testing:

# Use MCP dev command for automatic reloading
uv run mcp dev hue_server.py --log-level debug

# Or run directly with uv (stdio is default)
uv run python hue_server.py --log-level debug

Troubleshooting

• Bridge not found: If automatic discovery doesn't work, you have two options:

1. Manually edit the BRIDGE_IP variable in the script with your bridge's IP address
2. Manually create a config file:

     # Create the config directory
     mkdir -p ~/.hue-mcp
     
     # Create a config.json file with your bridge IP
     echo '{"bridge_ip": "192.168.1.x"}' > ~/.hue-mcp/config.json

Replace "192.168.1.x" with your actual Hue bridge IP address

• Connection issues: Delete ~/.hue-mcp/config.json and restart the server to re-authenticate
• Light control not working: Use refresh_lights tool to update the light information cache
• Groups or scenes not showing up: Restart the bridge and server to sync all data

How It Works

This server connects to your Philips Hue bridge using the phue Python library and exposes functionality through the Model Context Protocol. When an AI like Claude connects:

1. The server authenticates with your bridge using stored credentials
2. It provides resources that describe your lighting setup
3. It exposes tools that Claude can use to control your lights
4. It offers prompts that help Claude understand how to interact with your lights

All communication with your Hue system happens locally within your network for security and privacy.

Contributing

We are passionate about supporting contributors of all levels of experience and would love to see you get involved in the project. See the contributing guide to get started.

Project Structure

hue-mcp/
├── hue_server.py       # Main MCP server implementation
├── pyproject.toml      # Project configuration and dependencies
├── README.md           # This file
├── CONTRIBUTING.md     # Contribution guidelines
├── CHANGELOG.md        # Version history
├── LICENSE             # MIT License
├── tests/              # Test suite
│   ├── __init__.py
│   └── test_hue_server.py
└── .venv/              # Virtual environment (created during setup)

Development

# Install development dependencies
uv sync

# Run tests
uv run pytest

# Format code
uv run ruff check --fix hue_server.py

# Type check
uv run mypy hue_server.py

# Test with MCP Inspector
uv run mcp dev hue_server.py --log-level debug

License

This project is available under the MIT license. See LICENSE for details.