Hugging Face Datasets
AI & MLby huggingface
2/3 audits passApache-2.0
Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics.
Hugging Face Datasets, published by Hugging Face, teaches an agent read-only exploration of datasets through the Hugging Face Dataset Viewer API. The core workflow is: optionally validate a dataset with /is-valid, resolve its config and split with /splits, preview with /first-rows, paginate with /rows using offset and length, search text with /search, apply predicates with /filter, and fetch parquet links, size totals and column statistics with /parquet, /size and /statistics. It documents defaults such as the base URL, URL-encoding, zero-based offsets and the usual row limit per request, and notes that gated or private datasets need an HF token.
It also covers creating and uploading datasets, from the Hub web interface or the hub CLI, and uploading agent session traces as private datasets, since traces can contain prompts, file paths, secrets or personal data. Use it when you need to inspect, sample or pull data from a Hub dataset without downloading it all; it does not train models.
It is mainly for coding and data agents such as Claude Code. It is published under Apache-2.0 and importable into BusinessMCP, though it works with Hugging Face data rather than a workspace's own.
What you can do with it
- Preview the first rows of a Hub dataset split
- Filter a dataset's rows with a predicate
- Get parquet shard links and column statistics
- Upload agent session traces as a private dataset
Run it on your business data
Imported into BusinessMCP, Hugging Face Datasets becomes a playbook your AI business analyst applies to your connected tools and MCP servers.
Use Hugging Face Datasets in BusinessMCPInstall it in a coding agent
One command adds Hugging Face Datasets to your project.
npx skills add https://github.com/huggingface/skills --skill huggingface-datasetsHow we vetted it
- Source
- huggingface/skills at 80f9fa5
- Licence
- Apache-2.0
- Security audits (skills.sh)
- Gen Agent Trust Hub: Pass · Socket: Pass · Snyk: Warn
- Bundled scripts
- None, instructions only
Checked 2026-09-25 against its skills.sh listing. How we vet skills
Related skills
All skillsHugging Face LLM Trainer
Train or fine-tune language and vision models using TRL (Transformer Reinforcement Learning) or Unsloth with Hugging Face Jobs infrastructure.
Explore Data
Profile and explore a dataset to understand its shape, quality, and patterns. Use when encountering a new table or file, checking null rates and column distributions, spotting data quality issues like duplicates or…
Data Quality Frameworks
Implement data quality validation with Great Expectations, dbt tests, and data contracts. Use when building data quality pipelines, implementing validation rules, or establishing data contracts.
Frequently asked questions
Is the Dataset Viewer workflow read-only?
Yes. The exploration calls are GET requests, and the skill says to keep filtering and searches side-effect free. Separate sections cover uploading datasets.
Can it read private datasets?
Gated or private datasets require an Authorization header with your Hugging Face token.
How do I install Hugging Face Datasets?
Run `npx skills add huggingface/skills --skill huggingface-datasets`, or import it from the BusinessMCP dashboard.
Run Hugging Face Datasets against your whole business
Free plan, no credit card.
Use Hugging Face Datasets in BusinessMCP