hugging-face-datasets

Manage Hugging Face datasets with SQL queries via DuckDB and hf:// protocol.

3|Updated Feb 21, 2026

Search Tags:#data management #sql #data transformation #duckdb #datasets #huggingface

Ready to run

npx skills add https://github.com/ihatesea69/HieuNghi-AI-Skills --skill hugging-face-datasets-ihatesea69

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: hugging-face-datasets
Source: https://github.com/ihatesea69/HieuNghi-AI-Skills/tree/main/huggingface_skills/hugging-face-datasets
Command: npx skills add https://github.com/ihatesea69/HieuNghi-AI-Skills --skill hugging-face-datasets-ihatesea69

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill simplifies the creation, management, and querying of datasets on Hugging Face Hub, enabling efficient data manipulation and analysis directly through SQL.

Core Features & Use Cases

Dataset Lifecycle Management: Initialize new datasets, configure metadata, and stream updates.
SQL-Based Querying: Use DuckDB SQL to query, filter, transform, and aggregate data from any Hugging Face dataset using the hf:// protocol.
Multi-Format Support: Works with various dataset types including chat, text classification, QA, and tabular data.
Push to Hub: Save transformed data subsets as new repositories on Hugging Face Hub.
Use Case: You need to analyze a large dataset on Hugging Face, filter it based on specific criteria, aggregate some statistics, and then push the filtered subset to a new repository.

Quick Start

Query the 'cais/mmlu' dataset for nutrition-related entries using SQL.

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 using SQL?▼

You can query Hugging Face datasets using SQL by leveraging DuckDB and the hf:// protocol to filter, transform, and aggregate data directly without manual downloading. This approach supports various dataset types, including chat, text classification, and tabular data.

Can I push transformed data subsets back to the Hugging Face Hub?▼

Yes, you can push transformed data subsets back to the Hugging Face Hub. After filtering or aggregating data using SQL operations, the Skill enables saving the results as new repositories on the Hub for downstream access.

Does DuckDB support streaming updates for Hugging Face datasets?▼

DuckDB supports SQL-based querying and transformation for Hugging Face datasets. The Skill manages dataset initialization, metadata configuration, and streaming updates, allowing efficient data manipulation directly through SQL queries using the hf:// protocol.

What is the best way to manage large Hugging Face datasets for analysis?▼

The best way to manage large Hugging Face datasets is using SQL-based operations through DuckDB. This allows you to initialize datasets, stream updates, and perform SQL querying and transformation via the hf:// protocol to efficiently filter criteria and aggregate statistics.

What dataset formats work with the hf:// protocol in DuckDB?▼

The hf:// protocol in DuckDB works with multiple dataset formats, including chat, text classification, QA, and tabular data. This multi-format support enables versatile data handling and SQL-based querying across various Hugging Face dataset types.

Do I need Python scripts to transform Hugging Face datasets with SQL?▼

You need Python scripts to manage Hugging Face datasets through SQL-based operations. The Skill uses scripts alongside DuckDB and the hf:// protocol to execute dataset initialization, configuration, streaming updates, and SQL transformations.