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 qwen-delegategit 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/qwen-delegate)<a href="https://agentmods.dev/skills/sideshowroberto/vfx-agent-toolkit/qwen-delegate"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/qwen-delegate/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/qwen-delegate"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/qwen-delegate.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 | $0.00060 | $0.02178 |
| Opus 5.5 | $0.00024 | $0.00871 |
| Sonnet 5.5 | $0.00012 | $0.00436 |
| Haiku 4.5 | $0.00006 | $0.00218 |
Grade A, and why
qwen-delegate 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 — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Qwen Delegate
Routes a task to the local Qwen model served by Ollama (localhost by default).
Pointed at localhost, the prompt, any files the model reads, and the answer never
leave the machine, and the file content never enters Claude's context - only
the model's reply comes back. That reply CAN quote input content, so review what
returns before it lands in a hosted session. A remote --url or LOCAL_LLM_URL
sends the inputs off-machine: then it is not NDA-safe, whatever the model.
What you have installed
The team Qwen installer created ONE of these tags, chosen by GPU memory:
| Tag | Card | Context | Vision | Tools |
|---|---|---|---|---|
vfx-qwen38-27b-262k |
24 GB+ (workstation) | 262K | yes | yes |
vfx-qwen3-14b-16k |
under 24 GB | 16K | no | yes |
The scripts auto-detect whichever is present, so the commands below are the
same on every machine. Override with --model <tag> or the LOCAL_LLM_MODEL
env var. A different server (vLLM, llama-server, LM Studio) works via
--url / LOCAL_LLM_URL with --model local.
On the 14B variant: no --image (the script refuses with a clear message),
and one file read is capped at roughly 24K characters so it fits the 16K
context. Point the agent at a folder and ask about ONE file at a time.
When to delegate vs keep in Claude
| Delegate to Qwen | Keep in Claude |
|---|---|
| NDA-protected files (comp scripts, shot files, pipeline code on studio drives) | Multi-step reasoning, planning, architecture decisions |
| Summarizing large .nk / .py / .hip files | Tasks needing MCP tools (Nuke / Houdini / Unreal control) |
| Boilerplate and first-draft code | Complex debugging needing full conversation context |
| First-draft BlinkScript kernels | Editing files in the workspace |
| Text transformation, batch reformatting | Web search, library docs |
| Reference-image first pass (27B variant only) | Anything needing several MCP servers |
What ships with it
4 files 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.
- today First seen · 198 lines · 60 tokens per session scan A 91d9341fda6e
qwen-delegate is a skill published in the GitHub repository sideshowroberto/vfx-agent-toolkit (16 stars, last pushed 22d ago), licensed MIT. It adds 60 tokens to every session and 2,178 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
comfyui-civitai
A ComfyUI integration for searching, inspecting, and downloading models from CivitAI, a website that hosts community-created AI models and model versions.
flux-txt2img
Build Flux txt2img workflows with Flux.1 Dev (SRPO), Flux 2 Klein 9B, Turbo LoRAs, FluxGuidance, and DualCLIPLoader patterns.
prompt-engineering
ComfyUI prompt engineering knowledge covering CLIP text encoding syntax, weight modifiers, model-specific prompting strategies, and best practices.
civitai
Discover Civitai models with the BUILT-IN downloadmodel action:"searchcivitai" and install/generate them locally. Find a checkpoint/LoRA/embedding on Civitai, download it into ComfyUI, and use its trigger words. Optionally pair the official Civitai MCP for community features (images browsing, posting, collections).
scenario-model-training
Use when generated assets must keep a consistent style, character, or product look and references stop scaling, or when a user asks to train a custom model through the Scenario MCP, fine-tune a LoRA, clone a voice, curate a training dataset, choose a base model, set epochs and sample prompts, estimate training cost…
scenario-model-comparison
Use when comparing Scenario models on the same brief through MCP: a bake-off with shared prompts and inputs, cost, generation time and quality measured side by side, choosing a default model, testing a new release against the incumbent, or finding which model handles a style, edit, or reference image best. Keywords…