How do I install the Comfy MCP skill?

Run npx skillmds@latest add autonomous-ai/comfy-mcp in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Comfy MCP safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Comfy MCP?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Comfy MCP free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Comfy MCP?

autonomous-ai (@autonomous-ai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Comfy MCP by autonomous-ai — Agent Skill

Name: Comfy MCP
Author: autonomous-ai

File contents

Variation garden

Read studio.json to understand the current controls; "$STUDIO_TOOLCHAIN/../studio.config.json" describes their ranges. Run "$STUDIO_TOOLCHAIN/run.sh" generate to make a new result. Successful artifacts and their measurements are in out/runs/<id>/; out/latest.json names the current result. A failed run preserves the last success and records the error in the verdict.

The local starter creates deterministic procedural SVG artwork and a contact sheet. It is not diffusion or an AI-generated image claim. The native action submits the saved API workflow to an existing local ComfyUI, polls its result, and collects actual output images. The supplied native workflow uses no model and no paid API.

Use "$STUDIO_TOOLCHAIN/../README.md" for the integration contract and commands. Read the relevant files under $STUDIO_UPSTREAM before using an upstream API. Keep controls within their documented ranges, preserve the data needed to reproduce a comparison, and distinguish preview results from native service or hardware output. The viewer supports history and artifact downloads; tell the user which run contains the result, and what was actually measured.

Grow a reproducible family

Choose a palette and seed, run generate, then inspect each study and its saved recipe. The SVG hashes verify reproducibility. These local vector studies do not use a diffusion model. Keep selected studies and their seed/parameters together when making a new family.

For an installed ComfyUI, read $STUDIO_UPSTREAM/README.md and its src/ tools before using the MCP workflow. Save API-format nodes in workspace workflow.json. The supplied EmptyImage/SaveImage workflow needs no model weights. Run comfy against loopback COMFYUI_URL, inspect its job history, and retain the returned PNGs. A timeout means the remote queue may still be running; inspect it before resubmitting expensive work.

---name: comfy-mcpdescription: Create and inspect generative media projects in the Comfy MCP Harness workspace, including its local starter and optional upstream integration.---# Variation gardenRead `studio.json` to understand the current controls; `"$STUDIO_TOOLCHAIN/../studio.config.json"`describes their ranges. Run `"$STUDIO_TOOLCHAIN/run.sh" generate` to make a new result.Successful artifacts and their measurements are in `out/runs/<id>/`; `out/latest.json` names thecurrent result. A failed run preserves the last success and records the error in the verdict.The local starter creates deterministic procedural SVG artwork and a contact sheet. It is not diffusion or an AI-generated image claim. The native action submits the saved API workflow to an existing local ComfyUI, polls its result, and collects actual output images. The supplied native workflow uses no model and no paid API.Use `"$STUDIO_TOOLCHAIN/../README.md"` for the integration contract and commands. Read the relevantfiles under `$STUDIO_UPSTREAM` before using an upstream API. Keep controls within their documentedranges, preserve the data needed to reproduce a comparison, and distinguish preview results fromnative service or hardware output. The viewer supports history and artifact downloads; tell theuser which run contains the result, and what was actually measured.## Grow a reproducible familyChoose a palette and seed, run `generate`, then inspect each study and its saved recipe.The SVG hashes verify reproducibility. These local vector studies do not use a diffusionmodel. Keep selected studies and their seed/parameters together when making a new family.For an installed ComfyUI, read `$STUDIO_UPSTREAM/README.md` and its `src/` tools beforeusing the MCP workflow. Save API-format nodes in workspace `workflow.json`. The suppliedEmptyImage/SaveImage workflow needs no model weights. Run `comfy` against loopback`COMFYUI_URL`, inspect its job history, and retain the returned PNGs. A timeout means theremote queue may still be running; inspect it before resubmitting expensive work.

Frequently asked questions

Comfy MCP