pharma-skills

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides comprehensive educational and computational resources for pharmaceutical research without any malicious patterns.
  • [COMMAND_EXECUTION]: The drug-target-interaction skill invokes subprocess.run to call the vina binary. This is a standard practice for molecular docking simulations using the AutoDock Vina engine and does not allow for arbitrary command injection.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the chembl_webresource_client to retrieve drug bioactivity data from the ChEMBL database, which is a reputable and well-known service maintained by the European Bioinformatics Institute (EMBL-EBI).
  • [DATA_EXPOSURE]: Data processing is limited to local file handling for FAERS (FDA) data and structured API calls to public scientific repositories. No sensitive user credentials or private file access was detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 07:29 AM
Security Audit — agent-trust-hub — pharma-skills