huggingface-datasets

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the official @huggingface/hub utility and the parquetlens data tool via npx to facilitate dataset uploads and SQL-based querying.
  • [COMMAND_EXECUTION]: Employs standard system commands including curl, jq, and npx to fetch metadata from the Hugging Face Dataset Viewer API and execute dataset-specific tasks.
  • [DATA_EXFILTRATION]: Contains instructions for uploading local data folders to Hugging Face repositories using @huggingface/hub. This behavior is consistent with the skill's primary stated purpose of dataset creation and management.
  • [CREDENTIALS_UNSAFE]: Securely handles authentication by instructing the user to provide their Hugging Face token via an environment variable (HF_TOKEN) rather than hardcoding any secrets within the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 01:28 PM
Security Audit — agent-trust-hub — huggingface-datasets