pubmed-pico-search
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured methodology for performing clinical searches using Population, Intervention, Comparison, and Outcome (PICO) parameters. It utilizes specific, purpose-built tools (
parse_pico,unified_search,generate_search_queries) to translate user questions into structured search pipelines. No suspicious command execution, persistence mechanisms, or obfuscation techniques were detected. - [PROMPT_INJECTION]: The skill processes natural language clinical questions to extract structured data. While this represents a surface for indirect prompt injection where malicious instructions could be embedded in user queries, the skill provides a multi-step validation workflow that requires the agent to verify elements before execution, significantly reducing the risk of unintended behavior. The impact is limited to search query manipulation within trusted medical databases.
Audit Metadata