segment-modeling

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses industry-standard libraries including pandas, scikit-learn, and XGBoost to implement its core modeling functionality. No malicious behaviors such as data exfiltration or unauthorized network access were found.
  • [SAFE]: Data ingestion is limited to local file system access for Parquet and CSV files as specified by user arguments. The skill includes robust validation for file existence and supported formats.
  • [SAFE]: The implementation of rule-based segmentation uses pandas.DataFrame.eval() and pandas.DataFrame.query(). While these functions execute string expressions, they are used here in a legitimate context to allow users to define flexible data filtering rules, which is consistent with the skill's purpose as an analytical tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 06:40 AM
Security Audit — agent-trust-hub — segment-modeling