What problem does it solve?
This Skill eliminates the complexity of setting up and operating ComfyUI for generative media tasks, which normally requires manual server management, workflow JSON formatting, model dependency tracking, and API authentication for local or cloud execution.
Core Features & Use Cases
- Full Lifecycle Management: Install, launch, and manage ComfyUI servers and custom nodes using the official comfy-cli, with hardware-aware setup scripts that recommend local vs cloud deployment based on GPU VRAM.
- Workflow Execution & Parameter Injection: Run API-format ComfyUI workflows with dynamic parameter injection (prompts, seeds, dimensions), batch sweeps, img2img/inpainting support, and real-time WebSocket progress monitoring.
- Cross-Platform Cloud & Local Support: Seamlessly switch between local ComfyUI instances and Comfy Cloud hosted services, with automatic URL routing, auth header handling, and model list normalization across both environments.
Quick Start
Use the comfyui skill to generate an image by first running the hardware check script to determine if your machine should use local ComfyUI or Comfy Cloud, then execute a sample workflow with your custom prompt and save the resulting image to an output directory.