Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add huikku/comfyui-loop-mcp --skill docsgit clone --depth 1 https://github.com/huikku/comfyui-loop-mcpWrote 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/skills/huikku/comfyui-loop-mcp/docs)<a href="https://agentmods.dev/skills/huikku/comfyui-loop-mcp/docs"><img src="https://agentmods.dev/badge/skills/huikku/comfyui-loop-mcp/docs/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/skills/huikku/comfyui-loop-mcp/docs"><img src="https://agentmods.dev/badge/skills/huikku/comfyui-loop-mcp/docs.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 | $0.00116 | $0.03343 |
| Opus 5.5 | $0.00046 | $0.01337 |
| Sonnet 5.5 | $0.00023 | $0.00669 |
| Haiku 4.5 | $0.00012 | $0.00334 |
Grade C, and why
comfyui-workflows scanned grade C with 2 findings 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 1mo 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -s http://localhost:8188/object_info | python3 -c " Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
**If the [`comfyui-loop-mcp`](https://github.com/huikku/comfyui-loop-mcp) server is connected, use its tools instead of the curl below** — same method, less typing, and the parts curl can't do at all. The mapping: `check How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ComfyUI workflow generation
ComfyUI exposes a REST API that tells you exactly which nodes are installed, their inputs/outputs, and which model files exist. Always query the API before generating workflow JSON, and validate by executing before declaring a workflow done. Never guess node names, input names, types, or model filenames.
ComfyUI runs at http://localhost:8188 unless told otherwise.
If the comfyui-loop-mcp server is connected, use its tools instead of the curl below — same method, less typing, and the parts curl can't do at all. The mapping: check_comfyui (step 1) · list_nodes / get_node / list_models (steps 2-3) · check_workflow before you submit · submit_workflow + get_result (step 5) · get_image, which returns the pixels to you and compare_images(mode="difference") (step 6) · loop_start / loop_record to hold the best-so-far on disk, where a compacted context can't lose it. The curl in this skill is the fallback for when the server isn't there.
Step 0 — Pick the JSON format deliberately
Two formats, NOT interchangeable:
| API / prompt format | UI / litegraph format | |
|---|---|---|
| Shape | flat dict keyed by node id; each entry is class_type + inputs |
nodes + links arrays, positions, widgets_values |
| Used for | POSTing to /prompt to execute |
loading into the UI to view/edit |
| Generate when | "build it and run it" (default) | an artist needs to hand-tweak the graph |
Default to API format. It's robust to generate (no link-id accounting). Produce litegraph only when the goal is a UI-editable file. To run an existing UI workflow: convert it to API format (resolve reroute/GetNode/SetNode passthroughs and turn widgets_values into named inputs using object_info), validate in API format, then optionally save litegraph.
Install ComfyUI (only if it isn't already running)
Skip this whenever Step 1 succeeds.
If the comfyui-loop-mcp server is connected, load its comfy_install prompt and follow that — it is generated against the machine you are on: an install if one already exists (start it, don't clone a second), the interpreter to build the venv from, whether comfy-cli is on PATH, and the torch build this box's accelerator actually wants. Any failed tool call returns the short version of the same thing. Don't hand the install back to the user; you have the shell.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago First seen · 146 lines · 116 tokens per session scan C 70dd0e602f42
comfyui-workflows is a skill published in the GitHub repository huikku/comfyui-loop-mcp (3 stars, last pushed 2d ago), licensed MIT. It adds 116 tokens to every session and 3,343 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…