hugging-face-dataset-creator

Initialize and manage Hugging Face Hub datasets with template-based validation.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill streamlines creating and managing datasets on the Hugging Face Hub, including initializing repos, configuring system prompts, and streaming updates with validation templates.

Core Features & Use Cases

  • Initialize datasets: Create repos with proper structure and README metadata.
  • Configure & prompt: Store detailed configuration including system prompts and metadata.
  • Stream updates: Add data rows efficiently with template-based validation.
  • Template support: Work with chat, classification, QA, completion, and tabular data templates.

Quick Start

Examples:

  • Initialize a new dataset: python scripts/dataset_manager.py init --repo_id "your-username/dataset-name" [--private]
  • Add rows with validation: python scripts/dataset_manager.py add_rows --repo_id "your-username/dataset-name" --template qa --rows_json '[{"question": "Sample?", "answer": "Sample answer."}]'

Frequently Asked Questions about hugging-face-dataset-creator

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

FAQPage Schema
How do I create a dataset on the Hugging Face Hub with proper structure?

Initialize a dataset by running the dataset manager script with your repository ID. The initialization command creates repos with proper structure and README metadata, supporting private or public visibility on the Hugging Face Hub.

Can I add rows to my Hugging Face dataset with validation?

Yes, stream data rows with template-based validation using the add_rows command. Specify your template type—QA, classification, completion, or tabular—and provide rows as JSON; the system validates against the template schema before adding.

What dataset templates are supported for Hugging Face uploads?

The Skill supports chat, classification, QA, text completion, and tabular data templates. Each template enforces its own validation schema, enabling structured data entry across conversational, categorical, and structured formats.

Do I need to configure system prompts when setting up a Hugging Face dataset?

Configuration including system prompts and metadata is optional but recommended. Store detailed configuration during setup to enable consistent prompt behavior and metadata tracking across your dataset management workflow.

Can I manage multiple dataset formats in a single workflow?

Yes, the Skill handles multiple formats and templates in MCP-enabled workflows. Switch between QA, classification, text completion, and tabular data formats within the same repository, with each format validated against its schema.

How does batch update streaming work with Hugging Face datasets?

Stream row updates efficiently by providing JSON arrays of data objects to the add_rows command. The system processes batch updates with template validation, supporting iterative dataset building without manual formatting between additions.