Setup & Installation
openclaw skills install @dihan/comfy-ui
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install comfy-ui
What This Skill Does
Connects OpenClaw to a locally running ComfyUI instance for image generation. Uses an SDXL workflow by default (Juggernaut XL) and supports custom API JSON workflows. Images are saved locally and generation happens entirely on your own hardware.
All generation runs on your own machine, so prompts and images never leave your network and there are no per-image API costs.
When to use it
- Generating images privately without sending prompts to cloud services
- Running custom LoRA or fine-tuned model workflows from a home server
- Batch-generating concept art using saved workflow JSON files
- Automating image creation as part of a local agent pipeline
- Producing high-resolution SDXL outputs on dedicated GPU hardware
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User asks: Generate an image of a misty mountain at dawn
AGENT
- 1Reads COMFYUI_SERVER_ADDRESS from environment
- 2Calls comfy_gen.py with the prompt and server address
- 3Script connects to ComfyUI via WebSocket/HTTP and queues the prompt
- 4Waits for generation to complete and downloads the resulting image
- 5Saves image to image-gens/ folder and returns the file path
OUTPUT
Generated image saved locally, file path returned to user
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.
COMFYUI_SERVER_ADDRESS environment variable (e.g. http://192.168.1.119:8189)ComfyUI installed and running locally with Dev mode enabled