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 sideshowroberto/vfx-agent-toolkit --skill comfy-cli-skillsgit clone --depth 1 https://github.com/sideshowroberto/vfx-agent-toolkitWrote 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/sideshowroberto/vfx-agent-toolkit/comfy-cli-skills)<a href="https://agentmods.dev/skills/sideshowroberto/vfx-agent-toolkit/comfy-cli-skills"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/comfy-cli-skills/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/sideshowroberto/vfx-agent-toolkit/comfy-cli-skills"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/comfy-cli-skills.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 | $0.00058 | $0.01036 |
| Opus 5.5 | $0.00023 | $0.00414 |
| Sonnet 5.5 | $0.00012 | $0.00207 |
| Haiku 4.5 | $0.00006 | $0.00104 |
Grade A, and why
comfy-cli-skills scanned grade A with 0 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 today.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
comfy-cli skills
comfy-cli (Comfy-Org, GPL-3.0) ships its own agent skills for driving ComfyUI
from an agent: comfy (run and edit workflows), comfy-build, comfy-debug,
comfy-deploy, comfy-director (multi-shot narrative) and comfy-relay (media
presentation). They are not vendored in this MIT toolkit; install them from the
CLI so every harness reads the same text.
Install
pip install --upgrade comfy-cli
comfy --version
comfy skills install
comfy skills install needs comfy-cli 1.20 or newer; older builds expose the
same step inside the comfy setup wizard. Afterwards confirm the six comfy*
entries appear in the harness's skill catalog (restart the client if it caches).
The skills document a 1.20 command baseline: on an older CLI, read comfy --help
before using a command they name.
House rules around a PAID run
Partner nodes (Kling, Seedance, MiniMax, Veo, Gemini image) bill credits the moment a graph is accepted. Two zero-cost steps come first, every time:
- Validate, then print without posting.
comfy validate --workflow api.json --where localmust be clean, thencomfy run --workflow api.json --print-prompt --jsonlogs the exact graph. This catches wrong input names and DynamicCombo dot-path mistakes for free. "Clean" meansvalid=true,error_count=0and every warning READ, not merely counted. Known false positive (comfy validate 1.11.1): anedge_type_mismatchwarning when the expected input is a type UNION that already contains the produced type (e.g. "expects MESH,FILE_3D_GLB,...,FILE_3D but X produces FILE_3D_GLB") - the validator compares the whole comma-joined union as one string. That one warning is acceptable; any other warning stops the run. The validator can also be wrong the other way (pass a graph the server rejects), so the real pass is the expected output file on disk after the run, never the status. - Read the balance before and after.
GET https://api.comfy.org/customers/balancewith headerX-API-KEY: <key>;effective_balance_microsis US CENTS despite the name: value / 100 = USD, USD x 211 = credits (docs.comfy.org billing reference; comfy-cli PR #926). Reconcile the delta against the price badges. A larger drop than predicted means stop and find what else billed.
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.
- today First seen · 76 lines · 58 tokens per session scan A db6dda6fc6b3
comfy-cli-skills is a skill published in the GitHub repository sideshowroberto/vfx-agent-toolkit (16 stars, last pushed 22d ago), licensed MIT. It adds 58 tokens to every session and 1,036 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-02.
Other skills, from other repositories
artclaw-tool-executor
An instruction set for running tools from the ArtClaw Tool Manager, an interface for choosing and executing registered tools.
scenario-zbrush-expert
Use when an agent drives Maxon ZBrush 2026 for any task (sculpt in ZBrush, DynaMesh, ZRemesher, UVs, export, polypaint, posing, 3D print, clean up an AI mesh in ZBrush), when writing ZBrush Python (zbrush.commands) or ZScript, when the ZBrush bridge is not answering, a palette path or set() does nothing, a scripted…
scenario-zbrush-automation
Use when automating ZBrush 2026: batch processing a folder of OBJ or ZTL files, a ZBrush Python script or ZScript, recording a macro to learn a palette path, the ZBrush -script command line, scripting UV Master, Decimation Master or Multi Map Exporter, exporting every SubTool, GoZ or GoB round trips with Maya or…
dcc-cua
A routing guide for controlling application interfaces through the project’s DCC-CUA system. DCC applications are digital-content tools such as Maya, Blender, and Houdini.
mesh-ops
Cross-adapter polygon modeling recipe pack for the shared cross-DCC modeling vocabulary. Structured RECIPES.yaml entries for primitive creation, hard-surface edits, topology cleanup, UV generation, radial arraying, and material assignment. Reusable by every DCC adapter's mesh-ops skill (maya, blender, houdini, 3dsmax…
marketplace-publish-extension
Infrastructure skill — publish (register/update) an extension package to a marketplace catalog. Reads the extension's SKILL.md frontmatter, constructs a CatalogEntry, and upserts it into the target marketplace.json. Optionally commits and pushes when the catalog source is a git repository. Use after scaffolding an…