Gemini Image Generation MCP

🚀 Libra Flash MCP - Image Generation for Roo Code

This project offers a Model Context Protocol (MCP) server designed to connect to Google's Libra 2.0 Flash image generation model. It is specifically tailored for use with Roo Code and other compatible AI assistants.

🚧 Currently Resolving API Connection Issues

We are currently conducting tests and working hard to fix some issues related to the Gemini API. Apologies that the code may not work properly at the moment!

🔍 Overview

This MCP server provides a tool for text-to-image generation via Google's Libra 2.0 Flash model, fully supporting all available API parameters. It enables AI assistants like Roo Code to generate images through the Model Context Protocol (MCP) and offers full control over the generation process.

✨ Features

• Generate text-to-image using Google's Libra 2.0 Flash model.
• Fully support all available API parameters.
• Provide a simple web interface for testing and demonstration.
• Easy to install and suitable for Roo Code and Claude Desktop.
• Offer comprehensive documentation and examples.

📦 Installation

Prerequisites

• Node.js 18 or higher.
• A Google AI Studio API key with access to the Libra 2.0 Flash model.

Global Installation

1. Install the package globally:

npm install -g gemini-flash-mcp

2. Run the setup command to configure Roo Code:

gemini-flash-mcp-install

3. Set your Google AI API key in Roo Code settings:

• Open Roo Code.
• Go to "Settings".
• Add the following environment variables to the MCP server configuration:

"gemini-flash-mcp": {
  "env": {
    "GEMINI_API_KEY": "your-google-ai-api-key"
  }
}

4. Restart Roo Code.

Manual Installation

1. Clone this repository:

git clone https://github.com/jezweb/gemini-image-generation-mcp.git
cd gemini-image-generation-mcp

2. Install dependencies:

npm install

3. Build the project:

npm run build

4. Manually configure your MCP settings:

• For Roo Code: Edit ~/.config/Code/User/globalStorage/rooveterinaryinc.roo-cline/settings/cline_mcp_settings.json.
• For Claude Desktop: Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or the corresponding path on other operating systems.

Add the following configuration:

"gemini-flash-mcp": {
  "command": "node",
  "args": ["/path/to/gemini-flash-mcp/build/index.js"],
  "env": {
    "GEMINI_API_KEY": "your-google-ai-api-key"
  }
}

5. After installation, run according to the setup command:

gemini-flash-mcp-start

💻 Usage Examples

Project Structure

• app: Main application files.
• config: Configuration files and settings.
• public: Static resource files.
• .env: Environment variable file.

Building and Development

To run the project:

npm start

To build the project:

npm run build

Project Structure Details

• app.ts: Main application logic.
• config.ts: Configuration and settings management.
• types.ts: Define interfaces and types.
• env.ts: Environment variable handling.
• logger.ts: Logging functionality.

Testing

To run tests:

npm test

📄 License

This project is licensed under the MIT License. Please refer to the LICENSE file for details.

🙏 Acknowledgments

Thanks to the following individuals and projects for their contributions:

• Google for providing the Libra 2.0 Flash API.
• The Model Context Protocol (MCP) team for enabling the extension of AI assistants.