deepchem
Pass
Audited by Gen Agent Trust Hub on Apr 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides legitimate documentation and scripts for the DeepChem library, a standard tool in chemoinformatics and molecular machine learning.
- [EXTERNAL_DOWNLOADS]: The skill references and downloads datasets and pretrained models from well-known, trusted sources including the official DeepChem MoleculeNet repository and Hugging Face. These operations are transparent and consistent with the skill's stated purpose.
- [COMMAND_EXECUTION]: Python scripts utilize standard command-line argument parsing and DeepChem APIs for model training and evaluation. There is no evidence of arbitrary or dangerous command execution.
- [DATA_EXFILTRATION]: Data handling is performed locally or through established scientific repositories. There are no patterns suggesting unauthorized data exfiltration or exposure of sensitive user information.
- [PROMPT_INJECTION]: The instructions are technical and educational, focusing on ML workflows. No attempts were found to override agent instructions or bypass safety filters.
Audit Metadata