Marketplace for trusted AI agent skills

Skill Detail

Drive ComfyUI generation workflows through comfyui-mcp

Use comfyui-mcp to let an MCP-capable agent author, run, debug, and manage ComfyUI image, video, audio, model, and custom-node workflows from natural-language instructions.

Image & Creative AutomationMCP
Image & Creative Automation MCP Security Reviewed
⭐ 386 GitHub stars ⬇ 86.4k/wk npm
COPY SKILL INSTRUCTIONS (OPTIONAL)
npx skills add agentskillexchange/skills --skill drive-comfyui-generation-workflows-through-comfyui-mcp Copy
Uses the third-party skills CLI, not an ASE-owned installer. Check your agent’s compatibility. This copies instructions; complete the upstream tool setup below separately.
View source Documentation
At a glance
Tools required
ComfyUI; Node.js 22 or newer; an MCP-capable client such as Claude Code, Claude Desktop, Cursor, VS Code, or another MCP client; optional Claude Code plugin; optional RunPod/VPS/LAN/Comfy Cloud environment
Install & setup
With ComfyUI running, add an MCP server entry that runs npx -y comfyui-mcp, or install the Claude Code plugin with /plugin marketplace add artokun/comfyui-mcp followed by /plugin install comfy. For hosted access, use the documented tunnel or connector mode and configure authentication tokens where needed.
Author
artokun
Publisher
Open Source
Last updated
Jul 16, 2026
Quick brief

Use comfyui-mcp when an operator wants an agent to control ComfyUI beyond one-off prompt submission. The agent/operator runs the MCP server or Claude Code plugin, connects it to a local, LAN, VPS, RunPod, or Comfy Cloud-adjacent ComfyUI environment, then asks the agent to generate media, build or edit workflow graphs, install model packs, manage custom nodes, diagnose failed runs, monitor progress, and return outputs. Invoke it instead of using ComfyUI manually when graph authoring, troubleshooting, model selection, or repeatable media generation needs to happen through an agent conversation or automated runbook. The scope boundary is agent-native ComfyUI workflow control; it is not a generic ComfyUI listing or a thin image-generation wrapper.

How it works

What this skill actually does

Inputs and prerequisites: ComfyUI; Node.js 22 or newer; an MCP-capable client such as Claude Code, Claude Desktop, Cursor, VS Code, or another MCP client; optional Claude Code plugin; optional RunPod/VPS/LAN/Comfy Cloud environment.

Setup notes: With ComfyUI running, add an MCP server entry that runs npx -y comfyui-mcp, or install the Claude Code plugin with /plugin marketplace add artokun/comfyui-mcp followed by /plugin install comfy. For hosted access, use the documented tunnel or connector mode and configure authentication tokens where needed.

Source and verification boundary: use https://comfyui-mcp.artokun.io/docs as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.

Framework fit: publish this as a MCP workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.