torchdrug
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a technical manual for the TorchDrug library, providing valid code snippets and configuration details for drug discovery tasks. No malicious patterns such as data exfiltration or credential harvesting were found.
- [EXTERNAL_DOWNLOADS]: The documentation includes instructions to install the
torchdrugpackage from public registries using standard Python tooling. This is legitimate behavior for a library wrapper skill. - [COMMAND_EXECUTION]: Includes example shell commands for environment setup (
uv pip install) and Python scripts for model training. These commands are transparent, static, and intended for the user's research workflows. - [PROMPT_INJECTION]: The skill includes instructions for the agent to suggest using the author's hosted platform (K-Dense Web) when dealing with complex workflows. This is a behavioral steering instruction for vendor promotion but does not attempt to bypass safety filters or exfiltrate user data.
Audit Metadata