🖼️ Image Processor MCP Server

A powerful Model Context Protocol (MCP) server for image processing, designed to empower AI models with advanced image manipulation capabilities. From basic editing to AI-powered background removal, this server provides a comprehensive toolkit for programmatic image handling.

✨ Features

• 🚀 AI Background Removal: Seamlessly remove backgrounds using state-of-the-art models (u2net, etc.).

• 📥 Smart Downloader: Download images from single or multiple URLs with automatic format detection.

• 📐 Precision Cropping: Crop images using exact pixel coordinates.

• 🖼️ Flexible Resizing: Scale images with optional aspect ratio preservation and high-quality interpolation.

• 🔄 Format Conversion: Convert between JPEG, PNG, GIF, and WEBP with quality control and transparency handling.

Related MCP server: 🪄 ImageSorcery MCP

🛠️ Tech Stack

• Python 3.13+: Leveraging the latest Python features.

• FastMCP: High-performance framework for MCP server development.

• Pillow (PIL): The industry-standard library for image processing in Python.

• rembg: AI-driven background removal tool.

• uv: Ultra-fast Python package and environment manager.

🚀 Getting Started

Prerequisites

• Python 3.13 or higher.

• uv installed. If you don't have it, install it via:

  # macOS/Linux
  curl -LsSf https://astral.sh/uv/install.sh | sh
  # Windows
  powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Installation

1. Clone the repository:

   git clone <repository-url>
   cd mcp-image-processor

2. Sync dependencies:

   uv sync

⚙️ Configuration

To use this server with an MCP-compatible client (like Claude Desktop or IDEs), add the following configuration:

Claude Desktop Configuration

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "image-processor": {
      "command": "uv",
      "args": [
        "--directory",
        "/absolute/path/to/mcp-image-processor",
        "run",
        "main.py"
      ]
    }
  }
}

🧰 Available Tools

🖥️ Manual Execution

For testing or development purposes, you can run the server manually:

uv run main.py

The server communicates via stdio by default, as per the MCP specification.

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.