Frequently Asked Questions about ComfyUI MCP Server

Q: What is the ComfyUI MCP Server?

A: The ComfyUI MCP Server is a service that uses the Model Context Protocol (MCP) to enable image generation using natural language prompts with a local ComfyUI instance. It simplifies the process of creating images by abstracting away the complexities of ComfyUI’s node-based interface.

Q: What is MCP (Model Context Protocol)?

A: MCP is an open protocol that standardizes how applications provide context to Large Language Models (LLMs), facilitating better interaction and data exchange between AI models and external resources.

Q: What are the key features of the ComfyUI MCP Server?

A: Key features include natural language image generation, dynamic parameter adjustment, automatic workflow loading, MCP protocol compliance, and flexible deployment options.

Q: How do I install the ComfyUI MCP Server?

A: You can install the server using pip: pip install hh_mcp_comfyui. Ensure you have Python 3.12+ installed.

Q: How do I run the ComfyUI MCP Server?

A: After installing, run the server with the command: python -m hh_mcp_comfyui. Ensure your local ComfyUI instance is running.

Q: What is ComfyUI, and why is it required?

A: ComfyUI is a node-based visual programming environment for creating and executing Stable Diffusion workflows. The MCP Server interacts with a local ComfyUI instance to generate images.

Q: How do I configure the MCP server to work with ComfyUI?

A: Configure the COMFYUI_API_BASE environment variable to point to your ComfyUI instance (usually http://127.0.0.1:8188).

Q: Can I use the ComfyUI MCP Server with other clients like Cherry Studio or Cline?

A: Yes, the server is compatible with clients that support the Model Context Protocol (MCP). Configuration details are provided in the documentation.

Q: How do I add new workflows to the server?

A: Place your workflow JSON files in the designated workflows directory (e.g., /path/hh_mcp_comfyui/workflows). The server automatically loads these files on startup.

Q: What if I don’t have ComfyUI experience?

A: If you have no experience with ComfyUI, It is highly recommended that you first familiarize yourself with ComfyUI basic concepts by checking the official documentation. Also you can start by using the example workflows provided with the server and the 样例工作流插件安装教程.

Q: How do I test if the server is working correctly?

A: Use the MCP Inspector tool to test the server. Run the appropriate command for your deployment method (uvx, pip, or Docker) and connect to the server in the Inspector interface.

Q: What is UBOS and how does it relate to ComfyUI MCP Server?

A: UBOS is a full-stack AI Agent development platform that allows you to orchestrate AI Agents, connect them with your enterprise data, and build custom AI Agents with your LLM model and Multi-Agent Systems. The ComfyUI MCP Server can be integrated into UBOS to add image generation capabilities to your AI Agent workflows.

Q: Where can I find example workflows?

A: Example workflows are included in the example/workflows/ directory of the project repository.

Q: What should I do if I encounter issues or have questions?

A: You can reach out through the contact information provided in the documentation.