Skip to main content
affpapa.org
RU EN
skill Claude Code CX Codex Cursor GitHub Copilot

huggingface-datasets — Claude Code skill: what it does, how to install

Guides Dataset Viewer API work for inspecting Hugging Face datasets, rows, statistics, and parquet links. Use it when an agent needs read-only dataset exploration.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓ 
npx skills add huggingface/skills --skill huggingface-datasets -a claude-code
Open on GitHub

What huggingface-datasets does

This skill guides an agent through Dataset Viewer API requests. It can check whether a dataset is available, resolve configurations and splits, preview rows, page through content, search text, and apply row predicates. It also retrieves parquet links, size details, and column statistics.

Requests use GET, with URL-encoded query parameters. For row retrieval, the agent uses an offset and page length, then relies on response fields when a result is partial. Search targets string columns, while filtering accepts a predicate in where and an optional sort in orderby.

Gated or private datasets require Authorization: Bearer <HF_TOKEN>. Check the token before querying because those datasets cannot be read without it. The skill also covers parquet uploads and storing agent traces in private dataset repositories.

What it helps with

  • check whether a dataset can be viewed
  • list dataset configurations and splits
  • search dataset rows for text
  • filter records with a predicate
  • retrieve parquet shard links

Example prompts for the agent

  • “Check `<namespace/repo>` and list its configurations and splits with the Dataset Viewer API.”
  • “Read rows from `<namespace/repo>` for `<config>` and `<split>`, then continue through the remaining pages.”
  • “Search `<namespace/repo>` for `<text>` and filter rows using `<predicate>`.”

How to install huggingface-datasets

1Skills CLI (Vercel)

npx skills add huggingface/skills --skill huggingface-datasets -a claude-code

2Manually: put the files in place

~/.claude/skills/huggingface-datasets/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/huggingface-datasets/

From the same set10

The whole set →
huggingface-gradio huggingface This skill helps you build and modify Python Gradio interfaces for machine learning demos. It provides patterns for layouts, UI components, and chatbot workflows. Claude Code CX Codex Cursor +1 2,300installs huggingface-papers huggingface Pulls the paper ID out of a Hugging Face or arXiv link, reads the page as markdown, and fetches authors, models and datasets from the papers API. Claude Code CX Codex Cursor +1 2,200installs huggingface-llm-trainer huggingface Runs language and vision model fine-tuning on Hugging Face Jobs with TRL or Unsloth and converts checkpoints to GGUF without local GPUs. Claude Code CX Codex Cursor +1 2,200installs huggingface-community-evals huggingface Skill for running Hugging Face Hub model evaluations on local hardware via inspect-ai and lighteval. Use it to pick an inference backend or smoke-test a model before a full run. Claude Code CX Codex Cursor +1 1,900installs huggingface-trackio huggingface Tracks ML training experiments with Trackio, sets up error alerts, and queries metrics via CLI. Built for autonomous agent loops and Hugging Face Spaces. Claude Code CX Codex Cursor +1 1,900installs huggingface-local-models huggingface Finds a GGUF repo on Hugging Face, matches a quant to your memory and starts it with llama.cpp. No guessing at filenames. Claude Code CX Codex Cursor +1 1,900installs huggingface-paper-publisher huggingface Indexes arXiv papers on Hugging Face Hub, links them to model and dataset cards, verifies authorship and builds markdown drafts from templates. Claude Code CX Codex Cursor +1 1,700installs huggingface-tool-builder huggingface A skill for building reusable command-line scripts that fetch, enrich, and process data from Hugging Face. Useful when chaining multiple API calls or running repeated automation tasks. Claude Code CX Codex Cursor +1 1,600installs

Similar

arize-annotation official arize Creates annotation schemas and review queues in Arize, then updates span labels through the Python SDK. Use it for human review workflows. GitHub Copilot 1,100installs pathway-enrichment K-Dense-AI Runs pathway and gene-set enrichment analysis on gene lists or ranked data, finds over-represented GO terms, KEGG, Reactome, MSigDB. Needed after differential expression, CRISPR screens, or clustering. Claude Code CX Codex Cursor +1 1,400installs bulk-rnaseq K-Dense-AI Prepares bulk RNA-seq reads, Salmon, and STAR outputs into validated gene-level count matrices for differential expression in PyDESeq2. Claude Code CX Codex Cursor +1 1,400installs marimo-pair marimo-team Connects an agent to a live marimo notebook runtime to inspect variables and commit cell updates safely. Claude Code CX Codex Cursor +1 12,600installs dataset-curation Seth Hobson This skill formats and checks datasets for model fine-tuning. Use it to prepare data for SFT or preference training before a run. Claude Code CX Codex Cursor 2,600installs gmgn-market GMGNAI Fetches crypto candlestick charts, trending tokens, and launchpad feeds via GMGN API. Useful for tracking new meme coins and checking contract risks directly from the agent. Claude Code CX Codex Cursor +1 17,800installs xlsx official anthropics Creates, reads, edits, and converts spreadsheet files. Use it when the requested deliverable is an Excel workbook or another tabular file. Claude Code Claude.ai 187,223installs google-agents-cli-observability official Google This skill guides you through tracing and monitoring AI agents, including ADK agents. Use it when you need to inspect agent execution or monitor a deployed agent. Claude Code CX Codex Cursor +1 509,900installs

Questions about huggingface-datasets

How do I install huggingface-datasets in Claude Code?

Run npx skills add huggingface/skills --skill huggingface-datasets -a claude-code, then run npx skills add huggingface/skills --skill huggingface-datasets -a claude-code again. The skill file is at ~/.claude/skills/huggingface-datasets/SKILL.md.

What does huggingface-datasets do?

It guides Dataset Viewer API work for reading dataset metadata and rows, searching text, filtering records, retrieving statistics, and finding parquet links.

What is required for gated Hugging Face datasets?

Use the Authorization: Bearer <HF_TOKEN> header for gated or private datasets. The dataset exploration calls are read-only.

Source text

SKILL.md · Apache-2.0 
---
name: huggingface-datasets
description: 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 Dataset Viewer

Use this skill to execute read-only Dataset Viewer API calls for dataset exploration and extraction.

## Core workflow

1. Optionally validate dataset availability with `/is-valid`.
2. Resolve `config` + `split` with `/splits`.
3. Preview with `/first-rows`.
4. Paginate content with `/rows` using `offset` and `length` (max 100).
5. Use `/search` for text matching and `/filter` for row predicates.
6. Retrieve parquet links via `/parquet` and totals/metadata via `/size` and `/statistics`.

## Defaults

- Base URL: `https://datasets-server.huggingface.co`
- Default API method: `GET`
- Query params should be URL-encoded.
- `offset` is 0-based.
- `length` max is usually `100` for row-like endpoints.
- Gated/private datasets require `Authorization: Bearer <HF_TOKEN>`.

## Dataset Viewer

- `Validate dataset`: `/is-valid?dataset=<namespace/repo>`
- `List subsets and splits`: `/splits?dataset=<namespace/repo>`
- `Preview first rows`: `/first-rows?dataset=<namespace/repo>&config=<config>&split=<split>`
- `Paginate rows`: `/rows?dataset=<namespace/repo>&config=<config>&split=<split>&offset=<int>&length=<int>`
- `Search text`: `/search?dataset=<namespace/repo>&config=<config>&split=<split>&query=<text>&offset=<int>&length=<int>`
- `Filter with predicates`: `/filter?dataset=<namespace/repo>&config=<config>&split=<split>&where=<predicate>&orderby=<sort>&offset=<int>&length=<int>`
- `List parquet shards`: `/parquet?dataset=<namespace/repo>`
- `Get size totals`: `/size?dataset=<namespace/repo>`
- `Get column statistics`: `/statistics?dataset=<namespace/repo>&config=<config>&split=<split>`
- `Get Croissant metadata (if available)`: `/croissant?dataset=<namespace/repo>`

Pagination pattern:

```bash
curl "https://datasets-server.huggingface.co/rows?dataset=stanfordnlp/imdb&config=plain_text&split=train&offset=0&length=100"
curl "https://datasets-server.huggingface.co/rows?dataset=stanfordnlp/imdb&config=plain_text&split=train&offset=100&length=100"
```

When pagination is partial, use response fields such as `num_rows_total`, `num_rows_per_page`, and `partial` to drive continuation logic.

Search/filter notes:

- `/search` matches string columns (full-text style behavior is internal to the API).
- `/filter` requires predicate syntax in `where` and optional sort in `orderby`.
- Keep filtering and searches read-only and side-effect free.

For CLI-based parquet URL discovery or SQL, use the `hf-cli` skill with `hf datasets parquet` and `hf datasets sql`.

## Creating and Uploading Datasets

Use one of these flows depending on dependency constraints.

Zero local dependencies (Hub UI):

- Create dataset repo in browser: `https://huggingface.co/new-dataset`
- Upload parquet files in the repo "Files and versions" page.
- Verify shards appear in Dataset Viewer:

```bash
curl -s "https://datasets-server.huggingface.co/parquet?dataset=<namespace>/<repo>"
```

Low dependency CLI flow (`npx @huggingface/hub` / `hfjs`):

- Set auth token:

```bash
export HF_TOKEN=<your_hf_token>
```

- Upload parquet folder to a dataset repo (auto-creates repo if missing):

```bash
npx -y @huggingface/hub upload datasets/<namespace>/<repo> ./local/parquet-folder data
```

- Upload as private repo on creation:

```bash
npx -y @huggingface/hub upload datasets/<namespace>/<repo> ./local/parquet-folder data --private
```

After upload, call `/parquet` to discover `<config>/<split>/<shard>` values for querying with `@~parquet`.

## Agent Traces

The Hub supports raw agent session traces from Claude Code, Codex, and Pi Agent. Upload them to Hugging Face Datasets as original JSONL files and the Hub can auto-detect the trace format, tag the dataset as `Traces`, and enable the trace viewer for browsing sessions, turns, tool calls, and model responses. Common local session directories:

- Claude Code: `~/.claude/projects`
- Codex: `~/.codex/sessions`
- Pi: `~/.pi/agent/sessions`

Default to private dataset repos because traces can contain prompts, file paths, tool outputs, secrets, or PII. Preserve the raw `.jsonl` files and nest them by project/cwd instead of uploading every session at the dataset root.

```bash
hf repos create <namespace>/<repo> --type dataset --private --exist-ok
hf upload <namespace>/<repo> ~/.codex/sessions codex/<project-or-cwd> --type dataset
```

© huggingface · Apache-2.0 · huggingface/skills

See also

Project sponsors

Companies that keep this analytics open
Gold sponsor
Nutra.Club

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor