quality-assessment
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional, defining a subagent task for evaluating research rigor and reproducibility using a structured framework.
- [SAFE]: No external dependencies, package installations, or remote code execution patterns were identified.
- [SAFE]: No sensitive file access, credential usage, or network exfiltration patterns were detected in the prompt or configuration.
- [SAFE]: While the skill processes untrusted input in the form of paper details, it lacks any high-privilege capabilities or tools that could be abused through indirect prompt injection.
Audit Metadata