data-labeling

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured workflows and code templates for data labeling tasks using established tools like Label Studio and Snorkel.\n- [SAFE]: Credentials and sensitive configurations are handled securely through the use of descriptive placeholders (e.g., 'your-api-key').\n- [SAFE]: Network interactions are limited to a local environment (localhost), and file operations are restricted to expected dataset files (CSV, XML).\n- [SAFE]: Dependencies are restricted to reputable Python packages for data science and machine learning (pandas, numpy, scikit-learn, snorkel, label-studio-sdk).\n- [SAFE]: Data processing logic follows standard patterns for the task, with ingestion of local files for the purpose of annotation and quality control.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 07:16 AM
Security Audit — agent-trust-hub — data-labeling