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 newtongit 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/newton)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/newton"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/newton/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/newton"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/newton.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 | $0.00039 | $0.00333 |
| Opus 5.5 | $0.00016 | $0.00133 |
| Sonnet 5.5 | $0.00008 | $0.00067 |
| Haiku 4.5 | $0.00004 | $0.00033 |
Grade A, and why
newton 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 28d 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.
What it actually says
Newton
Newton is the physics engine tool for teacher policy training, demonstration generation, and policy evaluation in simulation.
The three core stages (train-teacher, generate-demos, eval) are
currently stubs: they expose the intended argument signatures but exit with
a clear "not yet implemented" message until the Newton simulation pipeline
lands (tracking issue nebius/nebius-physical-ai#499). Real physics validation
(a double-pendulum XPBD simulation) runs on CPU via the installed Newton
package — no GPU required for the validation path.
Interfaces
- CLI:
npa workbench newton <train-teacher|generate-demos|eval> --help - Python SDK (workbench-first):
npa.workbench.newton(train_teacher,generate_demos,evaluate) - Workflow module:
npa.workflows.byof.newton_pipeline(argument validation, stub-manifest plumbing, argparse entrypoint)
Conventions
- The CLI lives at
npa.cli.workbench.newton(multi-command package form for new tools); the SDK surface lives atnpa.workbench.newtonand does not go throughmake_cli_wrapper. - Stub stages write a
not_implementedmanifest to the requestedoutput_uribefore raising, so partial progress is always inspectable.
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.
- 28d ago First seen · 33 lines · 39 tokens per session scan A 3762c3c40de1
newton is a skill published in the GitHub repository nebius/nebius-physical-ai (50 stars, last pushed today), licensed Apache-2.0. It adds 39 tokens to every session and 333 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 2026-09-24.
Other skills, from other repositories
i4h-workflow-train-rl
Use when training, evaluating, or exporting Workflow policies with online RSL-RL or RLinf, including RL checkpoint and Workflow handoff.
godot-optimization
Use when optimizing Godot games — profiler, draw calls, physics tuning, memory management, and common bottlenecks.
metashape-reconstruction
Guide dense reconstruction, mesh building, and texturing in Metashape MCP. Covers depth maps, point cloud, mesh, texture, DEM, and orthomosaic generation with correct GPU/CPU settings and quality parameters. Works through the Metashape MCP server.
systemic-emergent-design
Use when designing, auditing, or generating the systemic substrate of a game — the small set of consistent, interacting rules (a "chemistry engine") that produce emergent play, rather than scripted set-pieces. Covers immersive-sim principles (intention & perceivable consequence), multiplicative vs. additive design…
gamebook-screenshot-analysis
Use to ingest Golf GameBook round screenshots (Round Summary + Statistics) into the coach. Extracts a coverage-aware round from the images, self-checks the read, stores it via the gamebook MCP tool (rolling last 5), and returns a normalized summary plus scoring-led progress vs prior rounds. Only score-per-hole is…
digital-twin-builder
Enriches a Universal Assembly Graph (UAG, produced by the game-designer skill) with simulation-grade semantics -- physical units, material/mass properties, and physics/logic constraints -- so it is ready for a physically meaningful digital twin, not just a visual scene. Use this skill whenever the user wants a UAG to…