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 agentmods add skills/graycodeai/starling/research-cosmos-policynpx skills add GrayCodeAI/starling --skill research-cosmos-policygit clone --depth 1 https://github.com/GrayCodeAI/starlingWrote 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/graycodeai/starling/research-cosmos-policy)<a href="https://agentmods.dev/skills/graycodeai/starling/research-cosmos-policy"><img src="https://agentmods.dev/badge/skills/graycodeai/starling/research-cosmos-policy.svg" alt="Measured on agentmods" 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 | $0.00040 | $0.00305 |
| Opus 5 | $0.00020 | $0.00152 |
| Sonnet 5 | $0.00008 | $0.00061 |
| Haiku 4.5 | $0.00004 | $0.00030 |
Grade A, and why
research-cosmos-policy 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 5d 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
Advanced topics
LIBERO command matrix: See references/libero-commands.md RoboCasa command matrix: See references/robocasa-commands.md
Resources
- Cosmos Policy repository: https://github.com/NVlabs/cosmos-policy
- LIBERO benchmark: https://github.com/Lifelong-Robot-Learning/LIBERO
- Cosmos-compatible RoboCasa fork: https://github.com/moojink/robocasa-cosmos-policy
- Upstream RoboCasa project: https://github.com/robocasa/robocasa
- MuJoCo documentation: https://mujoco.readthedocs.io/
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.
- 5d ago First seen · 36 lines · 40 tokens per session scan A 89a8acdc5ba1
research-cosmos-policy is a skill published in the GitHub repository GrayCodeAI/starling (2 stars, last pushed 6d ago), licensed MIT. It adds 40 tokens to every session and 305 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 2026-08-31.
Other skills, from other repositories
evaluating-cosmos-policy
Evaluates NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments. Use when setting up cosmos-policy for robot manipulation evaluation, running headless GPU evaluations with EGL rendering, or profiling inference latency on cluster or local GPU machines.
fine-tuning-serving-openpi
Fine-tune and serve Physical Intelligence OpenPI models (pi0, pi0-fast, pi0.5) using JAX or PyTorch backends for robot policy inference across ALOHA, DROID, and LIBERO environments. Use when adapting pi0 models to custom datasets, converting JAX checkpoints to PyTorch, running policy inference servers, or debugging…
fine-tuning-openvla-oft
Fine-tunes and evaluates OpenVLA-OFT and OpenVLA-OFT+ policies for robot action generation with continuous action heads, LoRA adaptation, and FiLM conditioning on LIBERO simulation and ALOHA real-world setups. Use when reproducing OpenVLA-OFT paper results, training custom VLA action heads (L1 or diffusion), deploying…
dfam-check
Measure mesh files against Design for Additive Manufacturing (DfAM) rules and report printability findings per process (FDM, SLS, SLA/DLP, metal PBF, MJF). Use when the user asks whether a part is printable, wants overhang/wall-thickness/support analysis of an .stl, .obj, .ply, or .3mf mesh, wants a build-orientation…
dart-io
DART IO: URDF, SDF, MJCF parsers, and dart::io loading.
dart-architecture
DART Architecture: the DART 7 multi-physics, multi-solver, multi-backend simulation pipeline and where each abstraction is owned.