Skills
Navigate
skills/google-deepmind/science-skills/clinical-trials-database/Gen Agent Trust Hub

clinical-trials-database

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from the ClinicalTrials.gov API, creating a surface for potential indirect prompt injection attacks.
  • Ingestion points: The scripts/clinical_trials_api.py script fetches study data, including summaries and eligibility criteria, from the external clinicaltrials.gov API.
  • Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded instructions within the trial data.
  • Capability inventory: The skill is capable of writing JSON data to local files using the --output flag in the scripts/clinical_trials_api.py script.
  • Sanitization: External text fields are not sanitized for instructions before being processed by the agent.
  • [EXTERNAL_DOWNLOADS]: The skill downloads the polite-http Python package via the uv package manager.
  • [COMMAND_EXECUTION]: The skill executes a custom Python wrapper script to perform API searches and data retrieval.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:47 PM