exia-mcp

# exia-scenario-generator MCP Server This is an MCP server for creating scenario files for the novel game engine [exia](https://github.com/kokushin/exia). It generates scenarios in the "Kotonoha Sisters Explanation" format and displays them in exia. <img width="800" alt="" src="https://github.com/user-attachments/assets/f02cc41a-6456-44d3-bb22-190cb899515d" /> ## Features - Generates scenarios in the "Kotonoha Sisters Explanation" format based on the specified theme - Converts the generated scenarios into JSON format for exia - Downloads and sets up exia from GitHub - Displays the generated scenarios in exia ## Requirements - Node.js v20.x or higher - OpenAI API key - Git ## Setup 1. Clone or download the repository ```bash git clone https://github.com/kokushin/exia-mcp.git cd exia-mcp ``` 2. Install the required packages ```bash npm install ``` 3. Compile TypeScript (run this after making changes) ```bash npm run build ``` ## How to Use with Claude Desktop 1. Launch Claude Desktop 2. Open the settings screen and add the MCP server ```json { "mcpServers": { "exia-scenario-generator": { "command": "npx", "args": [ "-y", "/path/to/exia-mcp", // (replace with the path to your cloned exia-mcp) "--openai-api-key=YOUR_API_KEY", // (replace with your OpenAI API key) "--stdio" ] } } } ``` - ※ Replace `YOUR_API_KEY` with your actual OpenAI API key - ※ If `npx` cannot be executed, specify the absolute path, e.g., `/path/to/.volta/bin/npx` 3. Use it in Claude Desktop as follows: - Input "Explain quantum computers in exia" - The MCP server will start, generate the scenario, and launch exia - Access exia in your browser to play the scenario ## Available Tools - `generateScenario`: Generates a scenario from a theme - `setupExia`: Downloads and sets up exia - `saveScenario`: Saves the generated scenario - `exiaVoiceroidExplain`: Executes the entire process from scenario generation to exia launch (recommended) ## Notes - The first execution may take time for downloading and setting up exia - Character images need to be prepared manually and replaced - The exia documentation can be found [here](https://github.com/kokushin/exia?tab=readme-ov-file#%E7%94%BB%E5%83%8F%E3%82%84%E3%82%B7%E3%83%8A%E3%83%AA%E3%82%AA%E3%82%92%E5%A4%89%E6%9B%B4%E3%81%97%E3%81%9F%E3%81%84%E5%A0%B4%E5%90%88) - Using the OpenAI API may incur costs ## License MIT