Skip to content
Get packs

ComfyUI Command-Line Control

Lets an agent queue ComfyUI image workflows, manage models, and download results via CLI

Source Download skill
Repository
HKUDS/CLI-Anything
Stars
58,084
Licence
Apache-2.0
Files
1 file

What it does

This skill teaches an AI agent to operate ComfyUI, a tool for generating AI images, through a command-line interface instead of its graphical window. Once installed, the agent can check the server, list available models, queue image-generation workflows, monitor progress, and download finished images. It is aimed at agents and power users who want to automate image generation.

Great for

  • Creators who already run ComfyUI locally and want to batch-generate images from scripts or an agent
  • Developers building automated image pipelines that need a scriptable ComfyUI interface with JSON output

Skip it if

  • People who do not have ComfyUI installed and running at localhost:8188, since the skill cannot work without it
  • Users who only want a visual, click-based interface, because this skill is entirely command-line

Try asking

  • “Check whether my ComfyUI server is running and show its GPU and memory stats”
  • “List the checkpoint and LoRA models available in my ComfyUI install”
  • “Queue my_workflow.json for generation and tell me when it finishes”
  • “Download all images from the last prompt run into ./outputs”
Step-by-step guide: how to use ComfyUI Command-Line ControlInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    cli-anything-comfyui.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip cli-anything-comfyui.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip cli-anything-comfyui.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Always pass --json when an agent calls the CLI, and verify the server with system stats before other commands
  • Use absolute paths for workflow and output files, as the SKILL.md recommends
  • Review workflow JSON before queueing, since it drives a local GPU and can consume significant resources

What's inside

  • SKILL.md

Redistributed unmodified from HKUDS/CLI-Anything under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.