huggingface-datasets

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [Command Execution]: The skill uses curl to interact with the Hugging Face Dataset Viewer API (datasets-server.huggingface.co). These are read-only GET requests used to retrieve dataset metadata and row previews.
  • [External Dependency Execution]: The skill utilizes npx to run parquetlens, @parquetlens/sql, and @huggingface/hub. These are well-known tools from the official Hugging Face ecosystem and the parquetlens project, used for querying parquet files and uploading datasets to the Hugging Face Hub.
  • [Data Exporting]: Instructions are provided for exporting dataset subsets to local files (CSV, JSON, Parquet) using SQL commands via parquetlens. This is a standard data processing task and occurs locally within the execution environment.
  • [Credential Management]: The skill references the use of an HF_TOKEN for accessing gated or private datasets. It correctly suggests using an environment variable (export HF_TOKEN=<your_hf_token>) for authentication, which is a standard practice for managing secrets in command-line environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 05:44 PM
Security Audit — agent-trust-hub — huggingface-datasets