infection-control-systems-expert
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override agent safety filters or bypass core instructions were detected. The content focuses on professional accountability and domain governance.
- [DATA_EXFILTRATION]: No hardcoded credentials, sensitive local file access, or unauthorized network communication patterns were identified.
- [EXTERNAL_DOWNLOADS]: No external package dependencies or remote script downloads were found in the instructions.
- [REMOTE_CODE_EXECUTION]: The skill does not contain patterns for executing remote code or dynamic runtime code generation.
- [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or persistence mechanisms were detected.
- [SAFE]: The skill defines a highly structured role with explicit rules for data validation, privacy impact assessments, and clinical safety reviews.
Audit Metadata