openevidence-reference-architecture

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation and reference code for a clinical decision support architecture. It follows standard software engineering practices for healthcare applications, including auditing, caching, and PHI sanitization.\n- [METADATA_POISONING]: Metadata fields such as name, description, and author are consistent with the skill's content and intended purpose, with no deceptive instructions detected.\n- [INDIRECT_PROMPT_INJECTION]: The architecture describes ingesting clinical queries from end-users, which represents an indirect prompt injection surface. However, the design includes sanitization steps (PHI handling) and the severity is minimal given the technical context of a reference architecture.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 10:05 PM
Security Audit — agent-trust-hub — openevidence-reference-architecture