ComfyUI
Run local ComfyUI workflows via the HTTP API. Use when the user asks to run ComfyUI, execute a workflow by file path/name, or supply raw API-format JSON; supports the default workflow bundled in assets.
Why use this skill?
Automate local ComfyUI image generation workflows with OpenClaw. Edit JSON prompts, seeds, and styles directly via your AI agent interface.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/kelvincai522/comfyuiWhat This Skill Does
The ComfyUI skill for OpenClaw empowers you to run sophisticated, local stable diffusion and generative image workflows directly from your agent interface. By interfacing with the ComfyUI HTTP API, the agent can execute complex, node-based image generation pipelines. The skill is designed to intelligently interpret user requests, map them to specific nodes within a JSON-formatted workflow, and automate the process of updating prompts, seeds, and style parameters before execution. It handles the heavy lifting of polling the local server and retrieving generated output images, making local AI art creation accessible via natural language commands.
Installation
To install this skill, use the standard OpenClaw installation command via your terminal: clawhub install openclaw/skills/skills/kelvincai522/comfyui. Ensure that you have a local instance of ComfyUI installed at ~/ComfyUI and that your system environment is properly configured to support the required Python virtual environment at ~/ComfyUI/venv/bin/python. If the server is not reachable, verify that the ComfyUI service is running on the default port (127.0.0.1:8188).
Use Cases
This skill is ideal for users who require precise control over their generative AI outputs. Use it for consistent character generation, applying complex upscaling processes, or running custom diffusion pipelines that aren't possible with basic image generators. It is perfect for professional workflows where you need to repeatedly modify parameters like CFG scale, sampler steps, or specific CLIP text encoder weights without manually opening the ComfyUI GUI. It is also excellent for batch-processing assets using a consistent workflow template.
Example Prompts
- "Run the default workflow and generate an image of a cyberpunk city street in the rain with neon lights."
- "Use the workflow at assets/portrait.json, set the seed to 12345, and generate a hyper-realistic portrait of an astronaut."
- "Execute my custom workflow and update the positive prompt node to include a cinematic lighting style with an 8k resolution tag."
Tips & Limitations
Always inspect your target JSON workflow before execution. Because ComfyUI workflows are node-based, node IDs can change if you add or remove nodes in the GUI, which may break the agent's ability to locate prompt or seed fields. If the agent cannot find a node, ensure the _meta.title or class_type names in your JSON match what the script expects. The skill performs file writes to temporary locations, so ensure your agent process has appropriate permissions in the ~/ComfyUI directory. If a workflow fails, check the console output for specific node errors, as mismatched model or checkpoint paths are common sources of failure in custom configurations.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-kelvincai522-comfyui": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read, code-execution, network-access