biomarker-landscape-scanner
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and uses reference frameworks to guide medical data analysis. It does not contain any executable code, obfuscation, or persistence mechanisms.
- [PROMPT_INJECTION]: The skill implements safety guardrails, including explicit out-of-scope redirects for patient-specific requests, which prevents the agent from being misused for clinical diagnosis.
- [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or unauthorized network operations were identified. The literature retrieval instructions point to standard biomedical databases.
Audit Metadata