zinc-database
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation and helper scripts facilitate the retrieval of chemical compound data and 3D molecular structures from official UCSF ZINC database domains, including cartblanche22.docking.org and files.docking.org.
- [COMMAND_EXECUTION]: The provided Python code snippets demonstrate the use of subprocess.run to invoke system-level CLI tools such as curl, wget, and aria2c for API communication and large-scale file transfers. These calls are implemented using list-based arguments which is a safe practice.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from an external chemical database, creating a surface for indirect prompt injection. * Ingestion points: Programmatic retrieval of substance data and SMILES strings from the CartBlanche22 API. * Boundary markers: None explicitly implemented in the example parsing code. * Capability inventory: Subprocess execution for network and file tasks; data manipulation via pandas; chemical property calculation via RDKit. * Sanitization: The provided code focuses on parsing structured TSV/CSV data and does not include specific sanitization logic for the content of the data fields.
Audit Metadata