design-experiment
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional, providing a procedural framework for scientific experimentation without executing code or scripts.
- [SAFE]: No evidence of prompt injection, obfuscation, or data exfiltration attempts was found in the instructions or metadata.
- [SAFE]: The skill does not perform network operations, access sensitive files, or install external dependencies.
- [SAFE]: The instructions focus on scientific best practices, such as pre-registration of metrics and derivation of sample sizes, and do not include any privilege escalation or persistence mechanisms.
Audit Metadata