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 nebius/nebius-physical-ai --skill wan2-2git clone --depth 1 https://github.com/nebius/nebius-physical-aiWrote 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/nebius/nebius-physical-ai/wan2-2)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/wan2-2"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/wan2-2/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/nebius/nebius-physical-ai/wan2-2"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/wan2-2.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00045 | $0.02240 |
| Opus 5 | $0.00023 | $0.01120 |
| Sonnet 5 | $0.00009 | $0.00448 |
| Haiku 4.5 | $0.00005 | $0.00224 |
Grade A, and why
wan2-2 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 4d 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.
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 — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Wan 2.2 Workbench support
Use this skill for the public Wan 2.2 registry candidate and its verified video evidence. Read these files before changing behavior:
workflows/testing/byof-wan2.2.yamlworkflows/testing/byof-wan2.2-multigpu.yamlnpa/src/npa/workflows/wan_rerun.pydocs/workbench/wan2.2.md
Also load byof-onboard, oss-solution-registry-onboard,
author-npa-workflow, real-components, solution-licensing, gpu-selection,
nebius-infra, testing-conventions, npa-agent, and
agent-visual-feedback when their surfaces are involved.
Ground truth
- Official source:
https://github.com/Wan-Video/Wan2.2.git, pinned to42bf4cfaa384bc21833865abc2f9e6c0e67233dc. - Official model:
Wan-AI/Wan2.2-TI2V-5B, pinned to921dbaf3f1674a56f47e83fb80a34bac8a8f203e. - TI2V-5B is a stock generative-video model supporting text and image inputs.
- A historical operator-only validation record accepted the real single-GPU
text-to-video path on RTX PRO 6000 Blackwell (
sm_120) from immutable image digestsha256:1baa4e2e89999ea26df81891ac786fa99c7498cbf173e5c5abad54c6f1dd1d13, including exact MP4/RRD byte identity. - A historical operator-only validation record accepted one shared official
generation from that same observed image digest on four B200s (
sm_100) with world size 4, NCCL, T5 and DiT FULL_SHARD FSDP, Ulysses size 4, and exact MP4/RRD byte identity. - Those records used Torch 2.7.1/CUDA 12.8 and NCCL 2.27.7. The current acceptance gate is Torch 2.13.0/CUDA 13.0 and NCCL 2.29.7; it requires fresh operator-accepted single- and four-GPU evidence before publication.
- I2V, A14B, speech-to-video, Animate, and training are separate capabilities.
- Stock Wan does not predict robot actions. Never claim that it is action-conditioned.
For changing facts, use only the official Wan repository, official Wan-AI model cards, and primary framework documentation.
Packaging contract
Use workbench.byof.repo; do not add a fake Wan toolRef. Keep the repo and all
model inputs immutable. The image may contain pinned source and dependencies but
no checkpoint weights, credentials, private code, or user data. The runtime
must remain non-root, with /opt/byof and its venv readable and 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.
- 4d ago Changed 8dafff3b3952
- 8d ago First seen · 168 lines · 45 tokens per session scan A cd4c36788304
wan2-2 is a skill published in the GitHub repository nebius/nebius-physical-ai (29 stars, last pushed today), licensed Apache-2.0. It adds 45 tokens to every session and 2,240 once invoked, about $0.0002 per session on Opus 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 2026-09-03.
Other skills, from other repositories
nvidia-speech-nim
Use NVIDIA Speech NIM microservices for speech production and voice workflows, including self-hosted ASR/STT, TTS, text translation, speech-to-speech pipeline design, Riva Python client integration, deployment planning, GPU/runtime sizing, licensing/privacy review, and production QA for transcription, captioning…
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
deploy-docker-compose
Run the Omnigent server as a Docker compose stack (server + Postgres) on any Docker host — your laptop, a VPS, EC2 by hand, or as the base layer of any container-platform deploy. Invoke when the user wants to build the image, bring up the compose stack, debug the stack on a host they already have, or extend the stack…