chem-nmr-analysis

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: All analysis scripts implement standard chemical data processing and visualization using NumPy, SciPy, and Matplotlib. No dangerous operations or obfuscated code were detected.
  • [EXTERNAL_DOWNLOADS]: The predict_products.py script communicates with HuggingFace's Inference API to predict reaction products. HuggingFace is a well-known and trusted service, and the data sent (SMILES strings) is sanitized and validated using RDKit prior to processing.
  • [COMMAND_EXECUTION]: Command usage is limited to standard CLI argument parsing. There are no calls to eval(), exec(), or subprocess modules with unsanitized user input.
  • [DATA_EXFILTRATION]: There is no evidence of the skill accessing sensitive files (such as SSH keys or AWS credentials) or unauthorized network transmission. Network activity is strictly confined to the stated functional purpose of reaction prediction via a trusted model provider.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 12:20 AM