Frequently Asked Questions (FAQ)

Q: What is an MCP Server? A: MCP (Model Context Protocol) is an open protocol standardizing how applications provide context to LLMs. An MCP server acts as a bridge, allowing AI models to access and interact with external data sources and tools.

Q: What types of assets can I generate with this server? A: You can generate both 2D and 3D game assets. For 2D, you can create pixel art, sprites, and textures. For 3D, you can generate models in OBJ and GLB formats.

Q: What is Hugging Face? A: Hugging Face is a company that provides an open-source platform and tools for natural language processing and machine learning, including pre-trained models, datasets, and APIs.

Q: Do I need a Hugging Face account to use this server? A: Yes, you need a Hugging Face account to access the AI models used by the server. You will need to obtain an API token from your Hugging Face settings.

Q: Which 3D model generation spaces are supported? A: The server supports InstantMesh, Hunyuan3D-2, and Hunyuan3D-2mini-Turbo. You will need to duplicate one of these spaces to your Hugging Face account.

Q: What is Claude Desktop? A: Claude Desktop is an example of an MCP client that can interact with the MCP server. It allows you to send commands and receive responses through a natural language interface.

Q: How do I install the server? A: You need to clone the repository from GitHub, install the dependencies using npm install, configure the .env file with your Hugging Face token and model space, and then run the server using npm start.

Q: What are the system requirements? A: You need Node.js version 16 or higher, Git, and internet access. You also need the required NPM packages listed in the documentation.

Q: How do I generate a 2D asset? A: Use the command generate_2d_asset prompt:"your prompt". Replace “your prompt” with a description of the asset you want to generate (e.g., generate_2d_asset prompt:"pixel art sword").

Q: How do I generate a 3D asset? A: Use the command generate_3d_asset prompt:"your prompt". Replace “your prompt” with a description of the asset you want to generate (e.g., generate_3d_asset prompt:"isometric 3D castle").

Q: Where are the generated assets stored? A: Assets are saved in the ./assets directory within the working directory of the server.

Q: How do I access the generated assets? A: Use the MCP URIs provided by the server (e.g., asset://2d_asset/filename.png) to list or read assets.

Q: Can I customize the asset generation process? A: Yes, you can customize various parameters related to 3D model generation, such as inference steps, guidance scale, and turbo mode, using environment variables.

Q: How do I troubleshoot errors? A: Check the server logs located in the ./logs/server.log file for detailed error messages. Also, verify your Hugging Face token and model space configuration.

Q: Is HTTPS supported? A: Yes, the server supports HTTPS for secure remote communication. You need to provide ssl/key.pem and ssl/cert.pem files.

Q: Where can I find more information about the UBOS platform? A: Visit the UBOS website at https://ubos.tech for more information about the platform and its features.