comfy-skills
What they did
Comfy's official plugin for coding agents. It adds slash commands and an MCP connection so an agent can generate images, video, audio and 3D, or run ComfyUI workflows on Comfy Cloud.
How it went
Installed as a Claude Code plugin; it needs a Comfy Cloud sign-in.
Worth knowing
- Price: Open source (MIT)
- License: MIT
- Status: sourced. We have not run this ourselves yet (no "Tested" badge).
- Quality bar (computed by code): 5/5 (primary source ✓, reproducible ✓, new ✓, example output ✓, safe ✓)
The best new AI video workflows and prompts, weekly.
Related
▶ demo at sourceComfyUI
ComfyUI is the open, node-based app for building image and video pipelines on your own GPU or in the cloud. Most open-model video workflows assume it.
▶ demo at sourceComfyUI docs: video
Official ComfyUI tutorial for Wan 2.1: text-to-video with the 1.3B model and image-to-video with the 14B model, including where each model file goes.
▶ demo on XComfyUI Changelog
ComfyUI's release log. Check it to see when new video models get native nodes.
▶ example at sourceComfyUI-MiniMaxH3-TimelineDirector
A ComfyUI plugin and all-in-one workflow for MiniMax H3. It covers text-to-video, reference images, audio-driven generation, editing and motion transfer, and builds long videos from continuous segments in one run.
▶ example at sourceComfyUI-MiniMaxH3-Easy
A single compact ComfyUI workflow for MiniMax H3 that covers text-to-video, image-to-video, first/last-frame and reference video, with one shared input panel.
▶ example at sourceComfyUI-Spectrum-MiniMax-H3
ComfyUI nodes that apply Spectrum, a training-free sampling speed-up from a 2026 paper, to the native MiniMax H3 audio-video model. They handle H3's combined audio/video latents.