duo-efficacy-measurement
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's content is entirely instructional and educational, focusing on psychometric evaluation and product measurement frameworks. No malicious patterns such as prompt injection, data exfiltration, or persistence mechanisms were detected.
- [SAFE]: External references are limited to official blog posts from a well-known service (Duolingo) for the purpose of citing methodology and research findings. These references are documented neutrally as legitimate educational sources.
- [SAFE]: There are no indicators of code execution, package installations, or privilege escalation. The skill does not utilize the dynamic context injection syntax (
!command).
Audit Metadata