personalization-logic
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Prompt Injection] (SAFE): No patterns indicating attempts to bypass safety filters or override system instructions were detected.
- [Data Exposure & Exfiltration] (SAFE): The skill does not access local files, system environment variables, or perform network requests.
- [Unverifiable Dependencies & RCE] (SAFE): No external dependencies, package managers, or remote script execution patterns were identified.
- [Indirect Prompt Injection] (SAFE): Although the framework discusses processing external data (segmentation), it contains no code or prompt interpolation logic that would create a vulnerability.
Audit Metadata