tooluniverse-drug-research
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from various external sources, including PubMed, ChEMBL, PubChem, and ClinicalTrials.gov. This information is used to progressively build a research report. The lack of explicit sanitization or instructions to ignore potential commands embedded in the retrieved scientific content creates a surface for indirect prompt injection.
- Ingestion points: Data is retrieved from external databases using tools such as
PubMed_search_articles,search_clinical_trials, andDailyMed_get_spl_sections_by_setidas defined in theSKILL.mdworkflow. - Boundary markers: The skill uses standard Markdown headers for report organization but does not include explicit delimiters or instructions to treat tool outputs as untrusted data.
- Capability inventory: The agent has the ability to create and write to local files (e.g.,
[DRUG]_drug_report.md) and utilize a wide array of specialized platform-provided tools. - Sanitization: There is no evidence of sanitization, escaping, or validation of the retrieved content before it is interpolated into the final report.
Audit Metadata