This MCP server implementation provides a bridge between ComfyUI and MCP clients for image generation tasks. It uses WebSockets for communication, allowing clients to send image generation requests with customizable parameters like prompt, width, and height. The server processes these requests through ComfyUI and returns the generated image URL. It's designed for easy integration into applications requiring on-demand AI image generation capabilities.
Aún no hay reseñas. ¡Sé el primero en reseñar!
Inicia sesión para unirte a la conversación
Highlight AIGenerate images (requires `prompt`)
Generate audio (requires `tags` and `lyrics`)
Regenerate an existing asset with optional parameter overrides (requires `asset_id`)
View generated images inline (images only, not audio/video)
Check ComfyUI queue state (running/pending jobs) - provides async awareness
Poll job completion status by prompt_id - check if a job has finished
Browse recently generated assets - enables AI memory and iteration
Get full provenance and parameters for an asset - includes workflow history
Cancel a queued or running job
List available ComfyUI models
Get current default values
Set default values (with optional persistence)
List all available workflows
Run any workflow with custom parameters
Show publish status (detected project root, publish dir, ComfyUI output root, and any missing setup)
Set ComfyUI output directory (recommended for Comfy Desktop / nonstandard installs; persisted across restarts)
Publish a generated asset into the project's web directory with deterministic compression (default 600KB)
