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/NVlabs/ASPIREnpx agentmods add skills/nvlabs/aspire/training-lawWrote 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/nvlabs/aspire/training-law)<a href="https://agentmods.dev/skills/nvlabs/aspire/training-law"><img src="https://agentmods.dev/badge/skills/nvlabs/aspire/training-law/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/nvlabs/aspire/training-law"><img src="https://agentmods.dev/badge/skills/nvlabs/aspire/training-law.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 | $0.00050 | $0.01765 |
| Opus 5 | $0.00025 | $0.00882 |
| Sonnet 5 | $0.00010 | $0.00353 |
| Haiku 4.5 | $0.00005 | $0.00177 |
Grade A, and why
robosuite/training-law/SKILL scanned grade A with 1 finding 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 9d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
echo "port $p: $(curl -s -o /dev/null -w '%{http_code}' --max-time 3 http://127.0.0.1:$p/health)" How it starts
The opening of the file, as written. The whole thing — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ASPIRE Experiment Pipeline
What Is ASPIRE
ASPIRE — LLMs write Python code to control a robot via a structured API. Code runs in MuJoCo (Robosuite).
7 Robosuite tasks:
| Task | Type | Camera key |
|---|---|---|
cube_lifting |
Single-arm | robot0_robotview |
cube_restack |
Single-arm | robot0_robotview |
cube_stack |
Single-arm | robot0_robotview |
nut_assembly |
Single-arm | robot0_robotview |
spill_wipe |
Single-arm | robot0_robotview |
two_arm_lift |
Bimanual | robot0_robotview |
two_arm_handover |
Bimanual | robot0_robotview |
Pipeline Modes
| Mode | File | When |
|---|---|---|
| Robosuite Baseline | run-baseline.md | Collect baseline on all 7 tasks |
| Robosuite Fix Loop (train-law) | main-agent-prompt.md | Coordinator guide: dispatch one subagent per GPU to debug failures |
| Robosuite Fix Loop Subagent (train-law) | subagent-prompt.md | Self-contained prompt template for dispatching one task to a background subagent |
Reference Files
| File | Covers |
|---|---|
| run-baseline.md | Launch command, config paths, output structure for all 7 tasks |
| api-reference.md | Full API functions, output structure, TraceLogger format, source files |
| clean-task-slate.md | Checklist for resetting a task to clean slate before rerunning fix loop |
Companion Skills
| Skill | Covers |
|---|---|
grasp |
Pick-and-place code template, pre-grasp/lower/close/lift/place skeleton |
localize |
Perception server (SAM3/Molmo) prompting strategy, per-object prompt registry |
transport |
Motion patterns for moving objects between locations — multi-step waypoints, safe transit sequences, interpolated Cartesian moves, collision avoidance during transport |
Setup
Two venvs:
.venv-robosuite(Python 3.10) — Robosuite replay/eval:replay_trial_robosuite.py.venv-liberoor.venv-perception— perception servers only (handled bystart_perception_servers.sh)
What ships with it
9 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.
- 9d ago First seen · 147 lines · 50 tokens per session scan A dd34fe0e22e6
robosuite/training-law/SKILL is a skill published in the GitHub repository NVlabs/ASPIRE (139 stars, last pushed 8d ago), licensed Apache-2.0. It adds 50 tokens to every session and 1,765 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
godot-optimization
Use when optimizing Godot games — profiler, draw calls, physics tuning, memory management, and common bottlenecks.
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…
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.
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…
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…
mosaic-scene
Max — Mosaic Bridge scene builder. Use for complete scene construction workflows — objects, materials, lighting, particles, physics, audio, and UI. Use when the user says "build a scene", "create a level", "add rain", "set up physics", or describes building any environment.