hugging-science

Discoveres and operationalizes curated scientific AI/ML resources from Hugging Face.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/gabrielvuksani/wotann --skill hugging-science
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hugging-science
Source: https://github.com/gabrielvuksani/wotann/tree/main/skills/scientific/hugging-science
Command: npx skills add https://github.com/gabrielvuksani/wotann --skill hugging-science

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Hugging Science solves the problem of wasting time on low-signal generic search when you need high-quality scientific datasets, models, and interactive demos for real AI/ML research work.

Core Features & Use Cases

  • Curated LLM-friendly catalog: Discover scientific datasets, models, and methodology posts across 17 domains via a curated index.
  • Domain-first discovery: Start from topic slugs (e.g., biology, chemistry, climate, scientific-reasoning) to narrow the search space quickly.
  • Actionable pointers to Hugging Face resources: Use catalog entries as launch points for loading datasets with datasets, running models with transformers, or calling interactive Spaces via gradio_client.
  • Workflow guidance for scientific specifics: Follow targeted reference docs for dataset loading, model execution (including trust_remote_code expectations), and Space API calling.
  • Practical citation support: When blog entries exist, use them to align methodology with the authors’ intended preprocessing and training choices.

Quick Start

Ask the agent to use hugging-science to find a suitable dataset and model for protein design, then fetch the relevant biology and biotechnology topic entries.

Frequently Asked Questions about hugging-science

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

FAQPage Schema
How do I find scientific datasets and models for machine learning research?

Finding scientific datasets and models requires parsing a curated Hugging Science catalog indexed by 17 domains, such as biology and physics, to select candidate entries by type and tags. You then use standard Hugging Face APIs to load the chosen resources.

How do I load and run scientific models from the Hugging Face Hub?

To load and run scientific models from the Hugging Face Hub, use the catalog entries as launch points for execution with transformers. Targeted reference docs provide workflow guidance, including handling trust_remote_code expectations for specific model implementations.

Can I use gradio_client to call interactive Spaces for scientific discovery?

Yes, you can use gradio_client to call interactive Spaces for scientific discovery. The catalog provides actionable pointers to Hugging Face resources, allowing you to use Space clients to execute chosen interactive demos directly.

What is the best way to locate AI resources for protein design?

The best way to locate AI resources for protein design is using domain-first discovery via topic slugs like biology and biotechnology. This narrows the search space quickly, yielding curated datasets and models specifically aligned with protein design tasks.

Does the Hugging Science catalog cover climate and materials science domains?

Yes, the Hugging Science catalog covers climate and materials science domains. It provides an LLM-friendly index across 17 domains to help researchers discover high-quality scientific datasets, models, and interactive demos for AI/ML work.

How do I align my preprocessing with a model author's intended methodology?

To align preprocessing with a model author's intended methodology, use practical citation support from blog entries linked in the catalog. These entries help align your methodology with the authors’ intended preprocessing and training choices.