Midjourney MCP

🚀 🎨 Midjourney MCP Server

A fully functional Midjourney MCP (Model Context Protocol) server that provides Midjourney image generation capabilities through the GPTNB API.

🚀 Quick Start

Environment Requirements

• Python 3.10+
• pip
• Node.js (for MCP Inspector, optional)

Installation

# Clone the project
git clone <repository-url>
cd mj-mcp

# Install dependencies
pip install -r requirements.txt

# Or install as an editable package
pip install -e .

Configuration

Set environment variables:

# Windows
set GPTNB_API_KEY=your_api_key_here
set GPTNB_BASE_URL=https://aiclound.vip

# Linux/Mac
export GPTNB_API_KEY=your_api_key_here
export GPTNB_BASE_URL=https://aiclound.vip

Or create a .env file:

GPTNB_API_KEY=your_api_key_here
GPTNB_BASE_URL=https://aiclound.vip

Running the Server

# Run directly
python src/server.py

# Or use the MCP development tool
npx @modelcontextprotocol/inspector python src/server.py

✨ Features

• 🖼️ Image Generation: Generate high-quality images via text prompts.
• 🔄 Image Transformation: Support operations such as enlargement, variation, and regeneration.
• 🎭 Advanced Editing: Editing functions like scaling, panning, and repairing.
• 👥 Face Replacement: Intelligent face swapping feature.
• 📝 Image Description: AI analysis of image content.
• 🔧 Prompt Optimization: Intelligent suggestions for optimizing prompt words.
• 📊 Task Management: Concise and efficient task status tracking.

🛠️ Available Tools (7 in total)

Core Image Generation Tools (6)

Task Management Tool (1)

📁 Project Structure

mj-mcp/
├── src/                    # Source code directory
│   ├── server.py          # Main entry of the MCP server
│   ├── client.py          # GPTNB API client
│   ├── service.py         # Business logic service
│   ├── config.py          # Configuration management
│   ├── models.py          # Data models
│   ├── task_handler.py    # Task handler
│   ├── exceptions.py      # Exception definitions
│   └── utils.py           # Utility functions
├── requirements.txt       # Project dependencies
├── setup.py              # Installation configuration
└── README.md             # Project documentation

🔧 Development

Code Structure

The project adopts a simplified flat structure, in line with the best practices of the MCP protocol:

• No package structure: Unnecessary __init__.py and __main__.py are removed.
• Direct entry: Start directly via python src/server.py.
• Modern dependencies: Use the latest version of the MCP SDK (1.9.1).
• Simple configuration: Manage dependencies using the traditional requirements.txt.
• Streamlined toolset: 7 core tools, focusing on practical functions and avoiding redundancy.

💻 Usage Examples

Basic Image Generation

# Call through the MCP client
result = await session.call_tool("imagine_image", {
    "prompt": "a beautiful sunset over mountains, digital art",
    "aspect_ratio": "16:9"
})

Image Variation

# Generate a variation
result = await session.call_tool("change_image", {
    "task_id": "previous_task_id",
    "action": "VARIATION",
    "index": 1
})

🤝 Contribution

Welcome to submit Issues and Pull Requests!

📄 License

GPL-3.0 License