healthcare-experiment-planning

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill defines an indirect prompt injection surface as it is designed to ingest and analyze user-provided healthcare experiment descriptions. Malicious instructions could be embedded within the data provided to the agent for critique.
  • [DATA_EXPOSURE]: The skill instructions include hardcoded absolute file paths (e.g., D:\code\agentic-memory\artifacts\...). While these refer to specific research documentation rather than sensitive system credentials, hardcoding absolute paths is a security anti-pattern that can expose project structure or enable unauthorized file access if the local environment matches the path schema.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 05:35 AM
Security Audit — agent-trust-hub — healthcare-experiment-planning