find-medical-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions require the agent to ingest and analyze untrusted user-generated content in the form of written patient reviews from multiple external platforms such as Google, Zocdoc, Healthgrades, WebMD, and Vitals.
- Ingestion points: External physician profiles and review pages accessed via browser tools during the search and screening phases defined in SKILL.md.
- Boundary markers: The instructions lack explicit security boundaries or guidance to the agent to disregard potential instructions embedded within the review text.
- Capability inventory: The agent is authorized to use browser tools for data gathering and spreadsheet tools for generating the final CSV output.
- Sanitization: There are no requirements for the agent to sanitize or escape the content retrieved from external review sources before processing or ranking.
Audit Metadata