Comfiest
A local-first, open-weights production suite for ComfyUI. Your agent runs image, video, speech, and music workflows on your own GPU.
Install / Use
claude mcp add BuffMcBigHuge -- npx -y github:BuffMcBigHuge/ComfiestIf the server publishes to npm under a different name, use that package instead β check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
AutomationSupported Platforms
Our assessment of Comfiest
Comfiest scores 84/100 on our quality scale, 1991st of 2,758 Automation skills we index.
Its MCP Server is 24 KB long, well organised into 19 sections with 5 code examples: a thorough specification that gives an agent plenty to work with.
It has 10 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated yesterday, so Comfiest is actively maintained.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 97/100, with no cautions. These come from repository metadata, not a code audit β read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands.
Automated pattern scan on 2026-11-21. It catches known dangerous patterns, not every risk β read a skill before letting an agent act on it.
Comfiest compared with similar skills
All 4 of these similar skills score higher than Comfiest; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| Comfiest (this skill)by BuffMcBigHuge | 84 | 10 | 1d ago | MCP Server |
| taste-skillby Leonxlnx | 100 | 120.0k | today | CLAUDE.md |
| Agent-Reachby Panniantong | 100 | 114.0k | today | CLAUDE.md |
| hyperframesby heygen-com | 100 | 88.2k | today | CLAUDE.md |
| rufloby ruvnet | 100 | 82.3k | today | CLAUDE.md |
Frequently asked questions
- How do I install Comfiest?
- Run
claude mcp add BuffMcBigHuge -- npx -y github:BuffMcBigHuge/Comfiest. The install tabs above show the steps for each supported agent. - Which AI agents does Comfiest work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is Comfiest safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. It is MIT-licensed and scores 97/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is Comfiest still maintained?
- The repository was last updated yesterday, so Comfiest is actively maintained.
Skill content
View source on GitHubNew in Comfiest
NEW β MiniMax H3 just leveled up. Comfiest ships the latest H3 innovations as runnable graphs and agent skills, not just links:
- Pika H3 Turbo LoRAs β the official Pika H3 Flex adapters from Pika-Labs/Pika-H3-Flex at the 6-step LCM contract, 24 fps, with text, first/last-frame, and reference modes. Pika T2V
- Upscale refiner LoRA β the Regenerate-2K refiner LoRA plus latent 2Γ upscale for 2560Γ1440 finals. Upscale workflows
- Fidelity 2K refiner β a 0.2 GB rank-21 Open-H3-Regenerate-2K-IC LoRA plus LMS Sharp, with no HyperFlow and no patching. It holds 32.4 dB PSNR to a Pika draft and keeps the draft audio bit-aligned. IC upscale
- 16 MP native image model β MiniMax-H3-Image: 33.1B, 6-step CFG 1, 4β16 MP natively, exact multi-line text, and 1/2/4-reference edit graphs (up to 16 supported). H3-Image T2I, edit, 2-ref, 4-ref
- Streaming 4Γ video upscale β ReCaVSR: one-step diffusion VSR on Wan2.2, 16.9 MP on 24 GB, with source audio and frame rate preserved. ReCaVSR X4
- Recast or edit any clip β Viggle Remake: a two-LoRA H3 pair for character swaps, object removal and weather/setting changes, driven by painted frames; source motion, camera and soundtrack are kept. Swap, edit
- Looping videos β seamless 154-frame H3 loops with synchronized audio and a Γ2 seam check, built on the MiniMax H3 base. Perfect loop
These are brand-new innovations in the MiniMax H3 ecosystem, and Comfiest supports all seven.
Comfiest packages the parts of an AI production pipeline that normally live in someone's head: model routing, workflow templates, hardware limits, submission loops, output provenance, and the small traps that waste a render. Point a coding agent at this repository and it can set up the MCP bridge, pick the right model for a shot, run it, and collect the result.
Built for the Comfy Org Open Source Higgsfield challenge.
| Create | Models and capabilities | | --- | --- | | Images from text | Krea 2 Turbo, Flux.2 Klein 9B, MOSS-Image, MiniMax-H3-Image, P-H3-Image | | Image edits | Krea 2 Edit, Flux.2 Klein 9B reference edit (1β4 references), MiniMax-H3-Image (1/2/4-ref graphs; up to 16 supported), P-H3-Image (up to 9 references) | | Speech | Fish Audio Drama 3 β single/dual speaker, voice clone | | Music | YuE2 3B, MiniMax Music 3 | | Video from text or references | MiniMax H3 β Pika H3 Flex 6-step default, fastvideo variants, 1β2 pictures, video/audio references, forced lip-sync, character swap | | Video recast / edit | Viggle Remake β H3 LoRA pair for character swaps, object removal, and setting/weather changes | | Video with joint audio | LTX-2.5 22B distilled β text, identity reference, or first frame | | Video upscale / restoration | RelayVSR X2 β one-step pixel-space 2Γ; ReCaVSR X4 β one-step streaming 4Γ with audio passthrough | | 2K upscale / refine | MiniMax H3 refiner β latent-space t2v/v2v upscale to 2560Γ1440, with 2-step re-synthesis or 3-step IC fidelity | | Seamless loops | MiniMax H3 Perfect Loop β text-to-looping-video with synchronized audio | | Mattes / segmentation | SAM3 β text-prompted image/video luma mattes, native ComfyUI nodes |
No partner nodes. No API keys. No credits. Everything runs on your own ComfyUI with open weights.
Start here: point your agent at the repo
The primary interface is an agent. Open this repository in Claude Code, OpenCode, Codex, Cursor, or any client with MCP support, and say:
βSet up Comfiest for my ComfyUI server, then run the fox demo and show me the output.β
The agent should:
- Read
README.md,COMFY-SETUP.md, andskills/SKILL.md. - If ComfyUI is not installed, follow
COMFY-SETUP.mdto install ComfyUI, the node packs, and the models. - Install/register the
comfy-*MCP server for your ComfyUI host usingMCP-SETUP.md. - Confirm the target with
server_info()and/system_stats. - Pick a workflow from the routing table and submit it with
run_workflow. - Collect the output into
work/outputs/<project>/<date>/<prompt_id>/and verify its size.
Add the skills to your agent
If your agent supports skills:
- Copy
skills/SKILL.mdto<client-skills>/comfiest/SKILL.mdfor the umbrella rules and routing. - Copy
skills/minimaxh3/,skills/minimaxh3-pika/,skills/minimaxh3-character-replace/,skills/minimaxh3-upscale/,skills/minimaxh3-perfect-loop/,skills/ltx2-5/,skills/relayvsr/,skills/ph3-image/,skills/h3-image/,skills/recavsr/,skills/viggle-remake/,skills/sam3/,skills/prompt-writing/,skills/workflow-development/, andskills/fish-audio-drama3/as sibling skill directories if your client discovers each skill separately.skills/prompt-writing/contains eight family skills; copy its child directories individually if your client requires a flat layout.
The umbrella carries the hard rules and routing; the subskills carry family and recipe detail.
What your agent gets
- Routing that respects hardware. Krea/Flux/MOSS/H3-Image for images, YuE2/Music 3 for songs, Drama 3 for speech, MiniMax H3/Pika for video, 2K refine, 4Γ upscale, recast/edit, and seamless loops. Heavy models are called out before they blow up a 24 GB card.
- Workflow templates that actually run. UI files for ComfyUI, pre-slimmed
api/*.api.jsonmirrors for MCP execution, and conversion notes for the frontend-only nodes that break API runs. - A development loop for new workflows. Scratch copies, UI-to-API conversion, target preflight,
submit/poll/collect, run-log entries, and a promotion checklist in
skills/workflow-development/SKILL.md. - A native matte primitive. SAM3 image/video segmentation with open-vocabulary prompts, coverage checks, and alpha-WebM compositing β no character-replacement graph required.
- A second video family. LTX-2.5 22B distilled generates video and audio in one pass from text, an identity reference, or a first frame.
- Prompt-writing skills for every family. Krea, Flux, MOSS, H3-Image, LTX, YuE2, Music 3, and Drama 3 prompts get the model-specific structure instead of a one-line caption.
- A submission loop that survives real life.
wait: false+/historypolling, direct output URL fallback when local job state lags, one-job-per-target discipline. - Provenance by default. Every file lands under a
prompt_idfolder with a run-log entry; vision QA is allowed only on your own outputs. - A writable workspace.
work/is the only place agents create files; the shipped workflows stay intact.
Quick start
1. Set up ComfyUI
Install ComfyUI 0.45 or newer with the node packs and models listed in
skills/SKILL.md. A 24 GB-class GPU is the baseline. Drama 3
loads 38.6 GB of experts into system RAM; use a high-RAM host.
For a complete, copy-pasteable install β ComfyUI, all 18 custom-node packs, every model, and the
TensorRT depth engine β follow COMFY-SETUP.md. It includes two scripts:
python scripts/install_nodes.py --comfyui "$COMFYUI"
python scripts/download_models.py --comfyui "$COMFYUI"
2. Connect the MCP bridge
Comfiest talks to ComfyUI through comfy-mcp, a thin stdio
server around comfy-cli.
# macOS / Linux
python -m venv ~/venvs/comfiest
~/venvs/comfiest/bin/pip install comfy-cli comfy-mcp
# Windows PowerShell
python -m venv $env:USERPROFILE\venvs\comfiest
& "$env:USERPROFILE\venvs\comfiest\Scripts\pip" install comfy-cli comfy-mcp
Then register one server per ComfyUI target. Full instructions, environment variables, and client
snippets for OpenCode, Claude Code/Desktop, Codex, Cursor, Hermes, and a generic stdio client are in
MCP-SETUP.md.
3. Run your first workflow
Ask the agent for the fox demo, or submit directly:
const run = await tools["comfy-<host>-<port>"].run_workflow({
workflow_path: "<repo>/workflows/text2image-krea.json",
wait: false
});
await tools["comfy-<host>-<port>"].job({ action: "wait", prompt_id: run.prompt_id });
await tools["comfy-<host>-<port>"].fetch_outputs({
prompt_id: run.prompt_id,
out_dir: "<repo>/work/outputs/fox-demo/<date>/" + run.prompt_id
});
Replace <repo> with the absolute path to this repository on the MCP host, and <host>/<port> with
one of your registered servers.
Demo inputs
The demo prompts are self-contained, but the reference-driven workflows expect files on the ComfyUI host:
| Workflow | Expected input | How to supply it |
| --- | --- | --- |
| Krea / Flux edit | example.png | Stage in work/uploads/<project>/, upload with upload_file. The Klein 2/3/4-ref graphs need every LoadImage slot set; all defaults are example.png. |
| Drama 3 clone variants | example.wav | Upload clean single-speaker clips and replace the transcript placeholders to match. |
| MiniMax H3 reference variants | example.png, example.mp4, example.wav | Upload the slots the variant uses; every image slot uses the same placeholder. |
| LTX-2.5 ref2v / i2v | example.png | Upload a subject image for ref2v, or a first frame for i2v; replace the file in work/uploads/<project>/. |
| RelayVSR | example.mp4 | Upload the source clip; output is 2Γ its size and keeps its audio. |
| ReCaVSR | example.mp4 | Upload the sourc
Truncated for display β read the full file on GitHub.
Related Skills
taste-skill
120.0kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Agent-Reach
114.0kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu β one CLI, zero API fees.
hyperframes
88.2kAI video you can actually edit. HyperFrames is a free, open-source AI video editor. Describe any video and get a real project you can change down to the frame.
ruflo
82.3kπ The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit β see the Safety scan above for what the skill file itself contains.
