physicsnemo-discover

Enumerate live PhysicsNeMo models, datapipes, and examples from repository files.

Updated Apr 16, 2026
One-click install
npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill physicsnemo-discover
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: physicsnemo-discover
Source: https://github.com/sayalinvidia/sayali-skills-test/tree/main/skills/physicsnemo-discover
Command: npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill physicsnemo-discover

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Discover and surface live PhysicsNeMo artifacts (models, datapipes, and examples) to support accurate, up-to-date architectural decisions and discovery workflows.

Core Features & Use Cases

  • Live-grounded discovery: enumerate current model families, datapipes, and documentation by querying the live repository state.
  • Flexible composition: surface orthogonal axes (model × datapipe × training strategy × config) and present concrete references to examples and docs.
  • Self-documentation: treat repository metadata (e.g., references/RECIPES.md, TAXONOMY.md) as the ground truth for navigation.
  • Use cases include selecting model families for a given data shape, choosing compatible datapipes, and locating relevant README/docs for implementation guidance.

Quick Start

Ask the skill to enumerate the live PhysicsNeMo model families and datapipes relevant to your data shape.

Frequently Asked Questions about physicsnemo-discover

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I discover PhysicsNeMo model families and datapipes for my data shape?

PhysicsNeMo discovery works by querying live repository files like SKILL.md frontmatter and references/RECIPES.md to surface current artifacts. It returns a small, orthogonal menu of candidates across up to four model families, three datapipes, two reference examples, and two docs that exist on disk.

What's the best way to select a compatible datapipe for a PhysicsNeMo model family?

The best way to select a compatible datapipe is to evaluate the orthogonal axes of model × datapipe × training strategy × config surfaced by the live repository. The skill presents concrete references to examples and docs to ensure the datapipe aligns with your chosen model family.

How does live repository metadata guide PhysicsNeMo architecture decisions?

Live repository metadata guides architecture decisions by treating files like references/RECIPES.md and TAXONOMY.md as ground truth for navigation. This ensures architectural choices rely on the current repository state rather than static information.

Can I locate PhysicsNeMo documentation and live examples for implementation guidance?

You can locate PhysicsNeMo documentation and live examples by querying the live repository state for files in the examples/* directory and references. The skill surfaces two reference examples and two docs that currently exist on disk for implementation guidance.

Do I need specific dependencies to enumerate live PhysicsNeMo artifacts?

No specific external dependencies are required to enumerate live PhysicsNeMo artifacts. The skill relies entirely on querying live repository files such as SKILL.md frontmatter, references/RECIPES.md, and examples/* to return current model families, datapipes, and documentation.