drugbank-database
Warn
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The documentation in 'references/data-access.md' provides examples using 'pickle.load()' to manage cached database results. This represents a dynamic execution risk (unsafe deserialization) as 'pickle' can be used to execute arbitrary code if the local cache files are tampered with.
- [PROMPT_INJECTION]: The skill is designed to ingest and process large quantities of drug information, such as descriptions and indications, which may contain malicious instructions. Ingestion points: Data is fetched from go.drugbank.com and processed via the helper scripts. Boundary markers: None identified; external data is integrated directly into the agent's context. Capability inventory: The skill has access to Python execution environment and network operations. Sanitization: No validation or escaping is applied to the text content extracted from the XML database.
- [EXTERNAL_DOWNLOADS]: The skill performs automated downloads of the DrugBank database from go.drugbank.com using the 'drugbank-downloader' package and 'requests' library.
Audit Metadata