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 cosmos3-npa-workflowgit 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/cosmos3-npa-workflow)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/cosmos3-npa-workflow"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/cosmos3-npa-workflow/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/cosmos3-npa-workflow"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/cosmos3-npa-workflow.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 | $0.00079 | $0.01741 |
| Opus 5 | $0.00039 | $0.00870 |
| Sonnet 5 | $0.00016 | $0.00348 |
| Haiku 4.5 | $0.00008 | $0.00174 |
Grade A, and why
cosmos3-npa-workflow 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 — 172 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cosmos 3 npa.workflow YAML
When To Use
Use this skill when the user wants a Cosmos 3 workflow they can submit with
npa workbench workflow submit — authoring the spec, validating it, or running
it. For the raw SkyPilot template, or for guardrail/prompt semantics, use
skills/workflows/cosmos3-inference/SKILL.md instead.
npa.workflow YAML Is The Supported Surface
The supported Cosmos 3 generation workflow is
workflows/testing/cosmos3-generate.yaml. The raw SkyPilot
template was retired after the spec twin reached a terminal live success in the
submit matrix, so do not recreate a SkyPilot task template for this path.
The npa.workflow spec resolves the container image, carries run-scoped S3 output URIs, and records declared outputs.
Reference Spec
For a complete source-video data factory, use workflows/main/paidf-cosmos3.yaml
and workflows/guides/paidf-cosmos3.md. Its metadata selects runtime execution
automatically, rejects assumed execution decisions, and enables source_overlay
so pinned images use the submitted NPA
adapters. Its structural_control=edge path prepares a complete 832×480 source
at conditioning_fps=24, drives native transfer in transfer_chunk_frames=93
windows with control_guidance=1.5, and verifies source/output timing and hashes
before evaluation. Model text and video guardrails remain enabled. Quality
acceptance uses the recorded grade_threshold and attribute_threshold;
lower thresholds never bypass media alignment, guardrails or endpoint failures.
Captioning and finalization require coverage of every accepted variant. Do not
describe generic prefix-conditioned video2video as equivalent to this path.
workflows/testing/cosmos3-generate.yaml is the working
example. Copy its shape:
apiVersion: npa.workflow/v0.0.1
kind: Workflow
metadata:
name: cosmos3-generate
config:
bucket: example-bucket
prefix: "runs/{{run.id}}/cosmos3-generate"
cosmos3_mode: text2image
cosmos3_checkpoint: Cosmos3-Nano
prompt: "a robot arm sorting colored blocks on a white workbench"
output_uri: "s3://{{config.bucket}}/{{config.prefix}}/generated/"
resources:
gpu:
cloud: kubernetes
accelerators: H100:1
cpus: 16
memory: 80Gi
initial: generate
states:
generate:
toolRef: workbench.cosmos3.generate
resources: gpu
outputs:
- uri: "{{config.output_uri}}generate.json"
schema: npa.cosmos3.generate.v1
terminal: true
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 Changed · +13 lines de9a8ef2ee78
- 4d ago Changed 4acd3e2fb77f
- 9d ago First seen · 159 lines · 79 tokens per session scan A 2d85a690f6f9
cosmos3-npa-workflow is a skill published in the GitHub repository nebius/nebius-physical-ai (29 stars, last pushed today), licensed Apache-2.0. It adds 79 tokens to every session and 1,741 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 2026-09-03.
Other skills, from other repositories
gke-compute-classes
Configures, optimizes, and troubleshoots GKE ComputeClasses. Use when configuring Spot VMs with on-demand fallback, targeting specific accelerators (GPUs/TPUs) or machine families, restricting ComputeClass access, or debugging pending pods related to node pool auto-creation. Do not use for cluster-level Node Auto…
gke-reliability
Improves GKE workload reliability, using PDBs, health probes, and topology spread constraints. Use when configuring GKE workload reliability, setting up PDBs, or configuring GKE health probes (liveness, readiness, startup). Don't use for disaster recovery setup or full cluster backups (use gke-backup-dr instead).
gke-workload-security
Audits, configures, and hardens workload-level security controls for Google Kubernetes Engine (GKE) applications and namespaces. Covers running cluster security audits (auditcluster.sh), configuring Workload Identity Federation (impersonation, KSA/GSA binding, and pod setup), enforcing Network Policies (default-deny…
nemo-automodel-launcher-config
Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.
azure-mgmt-botservice-dotnet
Azure Resource Manager SDK for Bot Service in .NET. Management plane operations for creating and managing Azure Bot resources, channels (Teams, DirectLine, Slack), and connection settings. Triggers: "Bot Service", "BotResource", "Azure Bot", "DirectLine channel", "Teams channel", "bot management .NET", "create bot".
cloud-architect
Designs cloud architectures, creates migration plans, generates cost optimization recommendations, and produces disaster recovery strategies across AWS, Azure, and GCP. Use when designing cloud architectures, planning migrations, or optimizing multi-cloud deployments. Invoke for Well-Architected Framework, cost…