pyhealth
Pass
Audited by Gen Agent Trust Hub on Apr 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses established healthcare machine learning workflows for processing Electronic Health Record (EHR) data. All code examples utilize the PyHealth library and standard PyTorch patterns for model training and evaluation.
- [EXTERNAL_DOWNLOADS]: The skill references official datasets from PhysioNet (MIMIC-III, MIMIC-IV, eICU) and the official documentation for the PyHealth project. These sources are recognized as reputable academic and technical repositories.
- [COMMAND_EXECUTION]: Includes standard package management commands for library installation using pip. No suspicious or high-risk command execution patterns were found.
- [DATA_EXFILTRATION]: File system operations are restricted to loading datasets from local paths provided by the user and saving structured model results (CSV) and embeddings (NPY) to the local directory. No unauthorized network transmissions or sensitive file access patterns were detected.
Audit Metadata