Skip to content
TakoAPI
Back to skills

dataset-finder

by anisafifi

Coding Agents & IDEs
1.2kDownloads reported by clawskills.sh
1Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

Use this skill when users need to search for datasets, download data files, or explore data repositories.

Description

What This Skill Does

Searches, downloads, and manages datasets from Kaggle, Hugging Face, UCI ML Repository, and Data.gov through a unified command-line interface. Supports format conversion, train/test splitting, dataset merging, and generating standardized documentation for any local dataset.

Searching Kaggle, Hugging Face, and UCI from one interface avoids switching between browser tabs and separate API clients for each repository.

When to Use It

  • Finding housing price datasets on Kaggle for a regression project
  • Downloading IMDB sentiment data from Hugging Face for NLP training
  • Previewing column types and missing value counts before committing to a large download
  • Generating a data card markdown file before sharing a dataset with teammates
  • Building an organized local dataset library across multiple ML projects

Example Workflow

INPUT

User asks: Find and download a sentiment analysis dataset from Hugging Face, then document it

AGENT
  1. 1

    Searches Hugging Face for datasets filtered by text-classification task and English language

  2. 2

    Downloads the selected dataset to the local datasets directory

  3. 3

    Runs preview on the downloaded files to check shape, column types, and missing values

  4. 4

    Generates a data card markdown file with schema, statistics, and usage examples

OUTPUT

Downloaded dataset in datasets/huggingface/imdb/ with DATACARD.md documenting schema, statistics, and sample usage

Requirements

  • Kaggle API credentials (kaggle.json) from kaggle.com/settings
  • Hugging Face account token from huggingface.co (required for gated datasets)
  • OpenClawCLI installed from clawhub.ai

Installation

clawhub install anisafifi/dataset-finder

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links