tooluniverse-clinical-trial-design

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes Python scripts (python_implementation.py and trial_pipeline.py) that implement an automated analysis workflow. These scripts are safe and perform data aggregation from established APIs.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with well-known biomedical and regulatory data services, including Open Targets, DrugBank, ClinicalTrials.gov, the FDA, and PubMed. These references are essential for the skill's primary research function.
  • [CREDENTIALS_UNSAFE]: The skill uses a .env.template file to guide users in setting up required API keys. This is a standard security best practice and does not contain any hardcoded or exposed secrets.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes clinical and research data from external databases and incorporates it into a final report. While this constitutes a data ingestion surface, the risk of indirect prompt injection is minimal given the specialized nature of the data sources and the structured report output format.
  • Ingestion points: Data from Open Targets, DrugBank, PubMed, and ClinicalTrials.gov.
  • Boundary markers: Data is encapsulated within specific sections and markdown formatting in the generated report.
  • Capability inventory: The skill performs read operations on external APIs and write operations to local markdown files.
  • Sanitization: No explicit sanitization of strings from external APIs is performed, as the data is primarily intended for human-readable reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 06:39 AM