#53 · Primary category: Image Generation

comfyui-mcp

agent-skills ai-agent claude-code claude-plugin comfyui comfyui-extension comfyui-mcp comfyui-mcp-server flux image-generation local-first mcp mcp-server model-context-protocol offline ollama self-hosted stable-diffusion video-generation wan

Local-first MCP server and sidebar agent that lets any LLM generate images, video, and audio, author workflows, and edit live ComfyUI graphs in natural language.

Project last updated:08/31/26

GitHub Stars

705

Forks

108

Contributors

20

License

MIT

Why we included this project

ComfyUI's node graphs are powerful but fiddly, and this MCP server is a way to let an LLM handle the wiring. It exposes a large toolset plus a sidebar agent that reads what is actually installed, builds or edits the graph node by node, frees VRAM, and runs the render, so a plain-language request for an image, video, or audio clip comes back as a finished file rather than a prompt handed to a thin relay. It works with whatever model you already pay for or run, including a fully offline Ollama setup with no account and no network, and one config covers a local machine, a LAN box, a VPS, or Comfy Cloud. The bundled model-specific skills carry curated sampler, CFG, resolution, and download guidance for architectures like Flux and WAN, which removes a lot of trial and error for people who want an agent that genuinely operates ComfyUI instead of just bridging to it.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category