Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/nebius/nebius-physical-ainpx agentmods add skills/nebius/nebius-physical-ai/alpamayo2-superWrote 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/alpamayo2-super)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/alpamayo2-super"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/alpamayo2-super/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/alpamayo2-super"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/alpamayo2-super.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 | $0.00074 | $0.01407 |
| Opus 5 | $0.00037 | $0.00704 |
| Sonnet 5 | $0.00015 | $0.00281 |
| Haiku 4.5 | $0.00007 | $0.00141 |
Grade A, and why
alpamayo2-super 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 3d 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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Alpamayo 2 Super
Run NVIDIA's real VLM plus 2.3B diffusion expert. Never replace inference with an import, fabricated trajectory, or manifest-only smoke.
Legal gates
Keep the boundaries separate:
- Source:
NVlabs/alpamayo2@beb2977d9a7e9d66837d4a3ad5144ff59de37519, Apache-2.0, baked with its license and the marked dataset-revision patch. - Model:
nvidia/Alpamayo2-Super@00554695e729a6ff0b6281fd2c81b18d06e33dbe, OpenMDW-1.1. Acceptance is by exercising granted rights. Fetch at runtime. - Dataset:
nvidia/PhysicalAI-Autonomous-Vehicles@b719eea7f0a63619ef51ec7f54178af0937ef050, gated by NVIDIA's AV Dataset License. It is non-transferable; accept it interactively on Hugging Face and fetch it only into the operator cache. - Outputs: OpenMDW-1.1 imposes no output restriction, but Alpamayo is not an automotive-grade driving stack. Preserve model-card safety provenance.
Before provisioning, run:
npa workbench health preflight
npa workbench health access --capability alpamayo2-super
A missing dataset entitlement is terminal. Open the exact URL printed by the access command; Hugging Face acceptance cannot be automated by NPA.
Build and scan
Build only from the checked-in Dockerfile. Never pass HF_TOKEN as a build arg
or populate /workspace/.cache/huggingface during a build.
bash npa/docker/workbench/alpamayo2-super/build.sh
npa/.venv/bin/python npa/scripts/scan_image_alpamayo2_payload.py \
<exact-local-image>
Require a complete clean scan over every layer for checkpoints, PhysicalAI-AV payload, caches, and credentials. Inspect SBOM/provenance separately; the byte scanner is not a license review.
Run the workflow
Use B200 first. The workload is headless and does not need RT cores; its measured
H100 peak is 72,115 MiB, so one B200 has ample memory. Validate RTX PRO 6000
separately because B200 sm_100 does not prove RTX sm_120.
npa workbench workflow validate-spec \
workflows/testing/alpamayo2-super-inference.yaml
npa workbench workflow submit \
workflows/testing/alpamayo2-super-inference.yaml \
--infra <configured-infra-target> --var bucket=<operator-bucket> \
--secret-env HF_TOKEN --secret-env AWS_ACCESS_KEY_ID \
--secret-env AWS_SECRET_ACCESS_KEY
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.
- 3d ago Changed · +10 lines 447baf9c2ebf
- 11d ago First seen · 111 lines · 74 tokens per session scan A 53a78c69c65e
alpamayo2-super is a skill published in the GitHub repository nebius/nebius-physical-ai (28 stars, last pushed today), licensed Apache-2.0. It adds 74 tokens to every session and 1,407 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
torchdrug
Build and troubleshoot TorchDrug 0.2.1 workflows for molecular graphs, property prediction, self-supervised pretraining, molecule generation, retrosynthesis, protein representation learning, and knowledge graph reasoning. Use when code imports torchdrug or needs its datasets, models, tasks, or Engine.
arboreto
Infer gene regulatory networks (GRNs) from gene expression data using scalable algorithms (GRNBoost2, GENIE3). Use when analyzing transcriptomics data (bulk RNA-seq, single-cell RNA-seq) to identify transcription factor-target gene relationships and regulatory interactions. Supports distributed computation for…
pyhealth
Build clinical/healthcare deep-learning pipelines with PyHealth — loading EHR/signal/imaging datasets (MIMIC-III/IV, eICU, OMOP, SleepEDF, ChestXray14, EHRShot), defining tasks (mortality, readmission, length-of-stay, drug recommendation, sleep staging, ICD coding, EEG events), instantiating models (Transformer…
deepspot-m
Generate transcriptome-wide virtual spatial transcriptomics from H&E histology with DeepSpot-M. Use when you need spatial gene expression in log1p-CPM for 224x224 tiles at about 20x, want to query protein-coding genes by symbol instead of a fixed panel, or want to run prediction across a whole slide after tiling with…
nemo-mbridge-perf-expert-parallel-overlap
Validate and use MoE expert-parallel communication overlap in Megatron-Bridge, including overlapmoeexpertparallelcomm, delaywgradcompute, and flex dispatcher backends such as DeepEP and HybridEP.
pick-a-pii-model
Select an on-device OpenMed PII model from the committed registry by language, runtime format, and size budget, then require recall validation before deployment. Use when an agent must choose a local PII detector for CPU, Apple Silicon, or a mobile export without relying on live model discovery.