artokun/comfyui-mcp

Local-first, agent-native control plane for ComfyUI — MCP server + sidebar agent that generates images, video & audio, authors and runs workflows, and edits your live graph in natural language on ANY LLM (Claude, ChatGPT, Gemini, offline Ollama, or any hosted model). 178 tools, 36 AI skills, 55 installer packs. Local, LAN, VPS, or Comfy Cloud.

What it solves

comfyui-mcp provides a comprehensive control plane for ComfyUI, allowing users to drive the complex node-based interface using natural language via any Large Language Model (LLM). It eliminates the need for manual graph wiring and trial-and-error parameter tuning by enabling an AI agent to author, edit, and execute workflows directly.

How it works

The project implements a Model Context Protocol (MCP) server that acts as a bridge between an LLM (such as Claude, ChatGPT, Gemini, or local models via Ollama) and a ComfyUI installation. It can operate as a standalone server, a Claude Code plugin, or as a sidebar agent panel embedded directly within the ComfyUI interface. The system uses a set of specialized tools and "skills" (model-specific guides for samplers, CFG, and resolutions) to ensure the agent can correctly configure workflows for different model architectures without manual intervention.

Who it’s for

It is designed for ComfyUI users who want to automate image, video, and audio generation, as well as developers who want to integrate ComfyUI's powerful generative capabilities into AI agent workflows across local, remote, or cloud-hosted environments.

Highlights

  • Multi-LLM Support: Compatible with Claude, ChatGPT, Gemini, and local models via Ollama.
  • Live Graph Editing: Ability to edit the live ComfyUI graph and spatial layout using natural language.
  • Extensive Toolset: Includes 38 MCP tools and 42 AI skills covering a wide range of models (Flux, WAN, LTX 2.3, etc.).
  • Autonomous Agents: Features specialized agents for exploring custom node packs, debugging workflow failures, and optimizing performance.
  • Local-First Architecture: Primary target is self-hosted ComfyUI, but supports remote installs via tunnels.
  • Integrated Panel: An autonomous AI agent sidebar that can be installed via ComfyUI-Manager.

Related

  • Project
    CoplayDev/unity-mcp

    MCP for Unity is an MIT‑licensed Unity package that connects LLMs (Claude, Codex, Gemini, local models, etc.) to the Unity Editor via the Model Context Protocol. It lets you create scenes, edit C# scripts, manage assets, run tests, and build games using natural‑language prompts from any MCP client. Install via Unity’s Package Manager, configure detected AI clients, and start issuing commands like “add a cube with a Rigidbody.” The project includes 50 ready‑made tool actions, multi‑instance routing, remote server support, and extensive documentation, with community channels on Discord and GitHub.

  • Project
    rusiaaman/wcgw

    An MCP server that gives AI agents unfiltered access to a local shell and file system for autonomous coding, building, and running of software.

  • Project
    isuzu-shiranui/UnityMCP

    A framework that exposes the Unity Editor as an MCP server, enabling AI agents to automate editor tasks, diagnose projects, and manipulate assets.

  • Project
    IvanMurzak/Unity-MCP

    Unity‑MCP is a Unity package that connects LLM agents (Claude, Codex, etc.) to the Unity Editor and to compiled games. It provides 70+ built‑in “tools” (asset management, scene manipulation, scripting, testing, profiling, etc.) that the AI can call, enabling natural‑language creation, debugging, and runtime AI behaviour. Install via a UnityPackage, OpenUPM, or the `unity-mcp-cli`; configure an agent, auto‑generate a skill set, and start issuing commands like “create three cubes in a circle”. Extensible with optional skill packs for Animation, Cinemachine, Navigation, and more. MIT‑licensed.

  • Project
    domdomegg/computer-use-mcp

    An MCP server that allows Claude and other AI models to control a local computer via keyboard and mouse automation.