hugging-face-datasets

Create, configure, and transform Hugging Face datasets with SQL via DuckDB.

Updated Feb 1, 2026

Search Tags:#machine learning #data engineering #sql #duckdb #datasets #huggingface

Ready to run

npx skills add https://github.com/nferrer-dev/claude-dotfiles --skill hugging-face-datasets-nferrer-dev

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: hugging-face-datasets
Source: https://github.com/nferrer-dev/claude-dotfiles/tree/main/skills/hugging-face-datasets
Command: npx skills add https://github.com/nferrer-dev/claude-dotfiles --skill hugging-face-datasets-nferrer-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires huggingface_hub, duckdb, datasets, pandas, and includes scripts (resource) components.

What problem does it solve?

This skill solves the friction of managing and analyzing large-scale machine learning datasets by providing a unified interface for creation, configuration, and SQL-based transformation.

Core Features & Use Cases

SQL-Based Data Wrangling: Use DuckDB to filter, aggregate, and transform datasets directly via the hf:// protocol without downloading entire files.
Dataset Lifecycle Management: Streamline the creation of new repositories, define system prompts, and manage dataset splits with built-in validation.
Use Case: Quickly create a high-quality training subset by querying a massive public dataset for specific criteria and pushing the filtered results to your own Hugging Face repository.

Quick Start

Use the hugging-face-datasets skill to query the cais/mmlu dataset and return the first ten rows where the subject is nutrition.

Frequently Asked Questions about hugging-face-datasets

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

FAQPage Schema

How do I query Hugging Face datasets with SQL without downloading the entire files?▼

You can query Hugging Face datasets with SQL using DuckDB via the hf:// protocol. This approach allows you to filter, aggregate, and transform data remotely without downloading entire files to your local environment.

How do I create and manage Hugging Face dataset repositories for machine learning pipelines?▼

You can manage Hugging Face dataset repositories by using tools that streamline dataset lifecycle management. This includes creating new repositories, defining system prompts, and managing dataset splits with built-in validation.

Can I use DuckDB to filter large public datasets and push the results to my own Hugging Face repository?▼

Yes, you can use DuckDB to filter large public datasets based on specific criteria. You can then push the filtered results to your own Hugging Face repository to create high-quality training subsets.

Does the hugging-face-datasets skill support schema discovery and data subsetting for ML training?▼

Yes, the hugging-face-datasets skill supports schema discovery and data subsetting for ML training. It utilizes the Hugging Face Hub API and DuckDB to enable efficient data manipulation and remote repository synchronization.

What is the best way to perform SQL-based transformations on Hugging Face datasets for data engineering?▼

The best way to perform SQL-based transformations on Hugging Face datasets is by using DuckDB. It provides a unified interface to directly query, filter, and aggregate data via the hf:// protocol for data engineering tasks.

What dependencies do I need to perform SQL-based data wrangling on Hugging Face datasets?▼

To perform SQL-based data wrangling on Hugging Face datasets, you need the huggingface_hub, duckdb, datasets, and pandas dependencies. These libraries facilitate remote querying, manipulation, and local data processing.