MCP Image Tools Server

A comprehensive MCP (Model Context Protocol) server that provides AI image generation and editing capabilities. This server integrates with Claude, Cursor, or other MCP-compatible AI assistants to offer powerful image manipulation tools.

Purpose

This server provides a complete suite of image tools for AI assistants, including:

• AI image generation using various models (FLUX, SDXL, Stable Diffusion, etc.)
• Image processing and editing (filters, cropping, resizing, rotation)
• Format conversion and metadata extraction
• Comprehensive model information and recommendations

Features

• Image Generation: Support for multiple AI models including FLUX, SDXL, ControlNet, and more
• Image Processing: Apply filters, crop, resize, rotate, and convert images
• Model Information: Get detailed information about available AI models, pricing, and performance
• TypeScript Support: Fully typed with proper type definitions
• Modular Architecture: Well-organized code structure with separate tool modules
• Easy Installation: Scripts for different MCP clients

Available Tools

Image Generation

• generate-image: Generate images using various Replicate models with extensive customization options

Image Processing

• apply-filter: Apply filters (blur, sharpen, brightness, contrast, etc.)
• crop-image: Crop images to specified dimensions
• resize-image: Resize or scale images with various fitting options
• rotate-image: Rotate images by specified angles
• convert-image-format: Convert between different image formats
• get-image-metadata: Extract detailed metadata from images

Model Information

• get-models: Get comprehensive information about available AI models including parameters, costs, and performance characteristics

Getting Started

# Clone the repository
git clone <your-repo-url>
cd mcp-images

# Install dependencies
pnpm install

# Set up environment variables
cp .env.example .env
# Edit .env and add your REPLICATE_API_TOKEN

# Build the project
pnpm run build

# Start the server
pnpm start

Installation Scripts

This server includes convenient installation scripts for different MCP clients:

# For Claude Desktop
pnpm run install-desktop

# For Cursor
pnpm run install-cursor

# For Claude Code
pnpm run install-code

# Generic installation
pnpm run install-server

These scripts will build the project and automatically update the appropriate configuration files.

Usage with Claude Desktop

The installation script will automatically add the configuration, but you can also manually add it to your claude_desktop_config.json file:

{
  "mcpServers": {
    "image-tools": {
      "command": "node",
      "args": ["/path/to/your/dist/index.js"],
      "env": {
        "REPLICATE_API_TOKEN": "your-replicate-api-token"
      }
    }
  }
}

Then restart Claude Desktop to connect to the server.

Usage Examples

Generating Images

# Generate an image using FLUX
generate-image --model "black-forest-labs/flux-schnell" --prompt "A beautiful sunset over mountains"

# Generate multiple images with different prompts
generate-image --model "stability-ai/sdxl" --prompt ["A cat", "A dog"] --num_outputs 2

# Generate with specific parameters
generate-image --model "black-forest-labs/flux-dev" --prompt "Cyberpunk cityscape" --width 1024 --height 1024 --guidance_scale 7.5

Processing Images

# Apply a blur filter
apply-filter --input_path "image.jpg" --filter "blur" --amount 5

# Crop an image
crop-image --input_path "image.jpg" --x 100 --y 100 --width 500 --height 500

# Resize an image
resize-image --input_path "image.jpg" --width 800 --height 600 --fit "cover"

# Convert format
convert-image-format --input_path "image.jpg" --format "png" --quality 90

Getting Model Information

# List all available models
get-models --category "all"

# Get information about a specific model
get-models --model_name "black-forest-labs/flux-schnell"

# Filter by category
get-models --category "flux" --include_pricing true

Environment Variables

Create a .env file in the root directory:

REPLICATE_API_TOKEN=your_replicate_api_token_here

You can get a Replicate API token from https://replicate.com/account/api-tokens

Project Structure

├── src/
│   ├── index.ts                    # Main server entry point
│   └── tools/                      # Tool implementations
│       ├── generate-image.ts       # Image generation tool
│       ├── image-processing.ts     # Image processing tools
│       └── model-info.ts          # Model information tool
├── scripts/                        # Installation and utility scripts
├── dist/                          # Compiled JavaScript (generated)
├── images/                        # Default directory for generated images
├── .env.example                   # Environment variable template
├── package.json                   # Project configuration
├── tsconfig.json                  # TypeScript configuration
└── README.md                     # This file

Development

1. Make changes to files in src/
2. Run pnpm run build to compile
3. Test your server with pnpm start
4. Use the installation scripts to update your MCP client configuration

Supported Models

The server supports a wide range of AI models including:

• FLUX Models: flux-schnell, flux-dev (fast, high-quality)
• SDXL Models: stability-ai/sdxl (high-resolution, detailed)
• Stable Diffusion: Various versions for different use cases
• ControlNet: For controlled generation with input images
• Specialized Models: Anime, realistic, artistic styles

Use the get-models tool to see detailed information about all available models, including pricing, performance metrics, and optimal use cases.

License

MIT