huggingface-datasets

Pass

Audited by Gen Agent Trust Hub on May 1, 2026

Risk Level: SAFE
Full Analysis
  • [Official API Interaction]: The skill uses the official datasets-server.huggingface.co API for read-only dataset exploration. This is standard functionality for the vendor and involves no unexpected network operations.
  • [Credential Management Recommendation]: The skill mentions the use of HF_TOKEN for accessing gated or private datasets. It correctly recommends using environment variables (export HF_TOKEN) rather than hardcoding credentials, which is a recognized security best practice.
  • [Verified Tool Usage]: The skill utilizes npx to run established packages such as @huggingface/hub and parquetlens. These tools originate from the official vendor or are widely recognized within the data science community for interacting with Parquet files and the Hugging Face Hub.
  • [Data Access Surfaces]: While the skill includes instructions for downloading data and exporting files (CSV, JSON, Parquet) using SQL queries via parquetlens, these operations are confined to the user's intended data exploration workflow and do not show signs of unauthorized data exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
May 1, 2026, 03:40 PM