sjtu-sai-agents/datamaster/dataset-search-tools

Claude Code and agent skill by sjtu-sai-agents/datamaster

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Dataset discovery and download toolkit covering HuggingFace, GitHub, Google, and academic paper search. Use when you need to find, inspect, or download ML datasets.

Install

npx vskill@latest install sjtu-sai-agents/datamaster/dataset-search-tools

SKILL.md

Dataset Search Tools

Search, inspect, and download datasets from HuggingFace Hub, GitHub, Google, and academic papers. All operations available as CLI scripts under scripts/. Source implementations copied from playground/search_dataset_tools/.

Download target: {workspace}/data_links/

HuggingFace

Primary source for ML datasets. HuggingFace search is keyword-exact, not semantic.

Key concepts: dataset_id (org/name), config (subset/language), split (train/validation/test).

ToolCLI
search_datasetspython scripts/hf_search.py search -q "sentiment" -l 10
search_datasets (author)python scripts/hf_search.py search -q "translation" -a Helsinki-NLP
inspect_datasetpython scripts/hf_search.py inspect --id stanfordnlp/sst2
inspect_dataset (config)python scripts/hf_search.py inspect --id stanfordnlp/sst2 -c default
get_dataset_configspython scripts/hf_search.py configs --id glue
get_dataset_splitspython scripts/hf_search.py splits --id stanfordnlp/sst2
get_dataset_samplepython scripts/hf_search.py sample --id stanfordnlp/sst2 -n 5 -s train
get_dataset_readmepython scripts/hf_search.py readme --id stanfordnlp/sst2
download_datasetpython scripts/hf_search.py download --id stanfordnlp/sst2 -o ./data_links/sst2

Typical workflow:

bash
python scripts/hf_search.py search -q "question answering" -l 10
python scripts/hf_search.py inspect --id openai/gsm8k
python scripts/hf_search.py sample --id openai/gsm8k -n 3
python scripts/hf_search.py download --id openai/gsm8k -o ./data_links/gsm8k

GitHub

Good for research datasets, benchmarks, curated collections.

ToolCLI
comprehensive_github_searchpython scripts/github_search.py comprehensive -k "benchmark dataset" --search-type all
search_repositoriespython scripts/github_search.py repos -k "nlp dataset" --sort stars
search_codepython scripts/github_search.py code -k "load_dataset" --language python
search_issuespython scripts/github_search.py issues -k "dataset release" --state open
search_pull_requestspython scripts/github_search.py prs -k "add dataset"
search_userspython scripts/github_search.py users -k "researcher"
get_repository_readmepython scripts/github_search.py readme --owner awesomedata --repo awesome-public-datasets

Web Search

Workflow: search first to get URLs, then parse specific URLs.

ToolCLI
google_searchpython scripts/web_search.py search -q "public NLP dataset download"
web_parsepython scripts/web_search.py parse -u "https://..."

Academic Papers

After finding PDF URLs, use web_parse to read content.

ToolCLI
arxiv_search_by_contentpython scripts/scholar_search.py arxiv -q "benchmark dataset" -n 10
arxiv_search_by_authorpython scripts/scholar_search.py arxiv-author -a "Yann LeCun" -n 5
google_scholar_searchpython scripts/scholar_search.py scholar -q "sentiment dataset"
search_dblp_paperspython scripts/scholar_search.py dblp-papers -q "Diffusion Model" -n 5
search_dblp_authorspython scripts/scholar_search.py dblp-authors -q "Geoffrey Hinton" -n 5
search_dblp_venuespython scripts/scholar_search.py dblp-venues -q "CVPR" -n 5

Data Validation

bash
python scripts/validate_dataset.py -p ./data_links/my_dataset

Checks file structure, CSV/JSON/text/parquet quality, missing values, duplicates, basic statistics.

Scripts Structure

scripts/
├── github_search.py
├── hf_search.py
├── scholar_search.py
├── src
│   ├── search_github.py
│   ├── search_huggingface.py
│   ├── search_scholar.py
│   └── search_web.py
└── web_search.py

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

See the full security report

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

See evaluation results

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/sjtu-sai-agents/datamaster/dataset-search-tools/badge)](https://verified-skill.com/skills/sjtu-sai-agents/datamaster/dataset-search-tools)

Something wrong with this skill or its page? Tell us.