Evaluating Cosmos Policy
This is a workflow for evaluating NVIDIA's Cosmos Policy model on robotics simulation benchmarks, specifically LIBERO and RoboCasa environments. It handles the full setup from scratch, including headless GPU evaluation and inference profiling. You'd reach for this if you're testing Cosmos Policy's performance or comparing it against other robotic manipulation models in standardized sim environments. The fact that it got 192 installs and passed most security audits suggests it's being used in actual research workflows. It's pulling from the public cosmos-policy repository, so you're working with the official eval setup rather than someone's fork.
- skills.sh installs
- 921
- GitHub stars
- 14.4K
- last updated
- Aug 21, 2026
Install
$ npx -y skills add orchestra-research/ai-research-skills --skill evaluating-cosmos-policy --agent claude-codeInstalls into .claude/skills of the current project.
Files
Cosmos Policy Evaluation
Evaluation workflows for NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments from the public cosmos-policy repository. Covers blank-machine setup, headless GPU evaluation, and inference profiling.
Quick start
Run a minimal LIBERO evaluation using the official public eval module:
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageDetails
- Repository
- github.com/orchestra-research/ai-research-skills
- Categories
- AI & Agent Building, Data Science & ML, Debugging
- First seen
- Jun 3, 2026
More from orchestra-research/ai-research-skills
All 95- Evolving Ai Agents897
- Ml Paper Writing2.5K
- Brainstorming Research Ideas1.3K
- Academic Plotting1.3K
- Creative Thinking For Research1.3K
- Mlflow1.1K
- Serving Llms Vllm1.1K
- Faiss1.1K
- Stable Diffusion Image Generation1.1K
- Tensorrt Llm1.1K
- Autoresearch1.1K
- Peft Fine Tuning1.1K
- Audiocraft Audio Generation1.1K
- Gguf Quantization1.1K
- Ray Train1.1K
- Weights And Biases1.1K
- Segment Anything Model1.1K
- Unsloth1K
- Crewai Multi Agent1K
- Evaluating Llms Harness1K
- Huggingface Accelerate1K
- Whisper1K
- Prompt Guard1K
- Knowledge Distillation1K
Recommended
More AI & Agent Building
anki-connectintellectronica/agent-skillsThis skill is for interacting with Anki through AnkiConnect, and should be used whenever a user asks to interact with Anki, including to read or modify decks, n918skills.sh installs
pptxailabs-393/ai-labs-claude-skillsPresentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or e916skills.sh installs
root-cause-analysisaj-geddes/useful-ai-promptsConduct systematic root cause analysis to identify underlying problems. Use structured methodologies to prevent recurring issues and drive improvements.916skills.sh installs
iso-13485-certificationk-dense-ai/scientific-agent-skillsPrepare and structurally review ISO 13485 QMS scope, controlled documentation, and local evidence manifests. Use for ISO 13485 certification-readiness documenta915skills.sh installs
pulumi-cdk-to-pulumipulumi/agent-skillsLoad this skill when a user wants to migrate, convert, port, translate, or move an AWS CDK application (including CDK stacks, constructs, or CloudFormation-synt912skills.sh installs