Borrowing it
Nothing to install: this file belongs to Comfy-Org/comfy-skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Comfy-Org/comfy-skills/main/AGENTS.mdgit clone --depth 1 https://github.com/Comfy-Org/comfy-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/comfy-org/comfy-skills/agents-md)<a href="https://agentmods.dev/instructions/comfy-org/comfy-skills/agents-md"><img src="https://agentmods.dev/badge/instructions/comfy-org/comfy-skills/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/comfy-org/comfy-skills/agents-md"><img src="https://agentmods.dev/badge/instructions/comfy-org/comfy-skills/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01024 | $0.01024 |
| Opus 5.5 | $0.00410 | $0.00410 |
| Sonnet 5.5 | $0.00205 | $0.00205 |
| Haiku 4.5 | $0.00102 | $0.00102 |
Grade A, and why
comfy-skills AGENTS.md scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl installer. Frozen; slated for removal once that installer retires. How it starts
The opening of the file, as written. The whole thing — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Comfy Cloud skills and plugins for AI coding agents: the slash commands and MCP connection that let an agent generate images, video, audio, and 3D, search models/nodes/templates, and run ComfyUI workflows through the hosted Comfy Cloud service. Everything here is plain Markdown (command/skill files plus JSON manifests) — there is no application code, build step, or compiled artifact.
Commands
This repo has no package manifest, test suite, or lint config. The one check the project runs is Claude Code's plugin-manifest validator (README "Contributing"):
claude plugin validate ./claude-code # validate the comfy-cloud plugin before pushing
Run it from the repo root; it validates claude-code/.claude-plugin/plugin.json.
There is nothing to install, build, or run locally — the plugin points agents at
the hosted MCP server (https://cloud.comfy.org/mcp), it does not run a server.
Directory map
claude-code/— thecomfy-cloudClaude Code plugin. Edit commands here.claude-code/commands/— the slash-command Markdown files (generate-image,generate-video,search-templates,help, …).claude-code/.claude-plugin/plugin.json— plugin metadata + the bundledcomfy-cloudHTTP MCP server declaration.
.claude-plugin/marketplace.json— marketplace manifest so/plugin marketplace add Comfy-Org/comfy-skillsresolves the plugin.skills/— legacy flat command files for the deprecatedcomfy-cloud-mcpcurl installer. Frozen; slated for removal once that installer retires.openclaw/comfy/SKILL.md— the OpenClaw-format skill for the hosted MCP.assets/— static assets (the README logo)..github/workflows/— CI (see below).
Conventions and gotchas
- Authoring rule — steer the approach, defer the specifics (README has the
full version). A command is frozen on the user's machine at install time, so it
must never hard-code facts that rot: template names, live model names, node ids.
Write stable steering (the approach) and tell the agent to discover current
options at run time via the tools (
search_templateswithtag,search_nodeswith typed params) instead of pasting today's catalog inline. If you catch yourself typing a literal model/template/node name into a command, point the agent at the tool that returns the current set instead. - Commands are thin Markdown with a short YAML frontmatter
description. Add or edit underclaude-code/commands/, keep prose clear, and avoid emoji. claude-code/is canonical;skills/is frozen. Prefer editing the plugin underclaude-code/. Don't build new work on the legacyskills/layout.- comfy-cli skills live elsewhere. The
comfy,comfy-debug,comfy-build, etc. agent skills are no longer in this repo — they ship bundled inside comfy-cli (comfy_cli/skills/) and install viacomfy skills install. Don't re-add them here. - Commit style — Conventional Commits with an optional scope, matching the
git history:
feat(generate-video): …,fix(generate-audio): …,chore: …,docs(openclaw): …. Land changes via pull request. - Public repo. This repository is public; keep everything committed here fit for a world-readable audience (no secrets, no private internal detail).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 77 lines · 1,024 tokens per session scan A 5af7d878e068
comfy-skills AGENTS.md is an instructions file published in the GitHub repository Comfy-Org/comfy-skills (236 stars, last pushed 2d ago), licensed MIT. It adds 1,024 tokens to every session, about $0.0041 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-23.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.