counterfactual-generator
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured instructions for scientific reasoning and counterfactual analysis. It contains no executable scripts, shell commands, or network activity indicators.
- [PROMPT_INJECTION]: No attempts to override safety filters or bypass core instructions were identified. Indirect injection assessment: Ingestion point is the literature corpus; boundary markers are absent; no executable capabilities detected; sanitization is absent.
- [DATA_EXFILTRATION]: There are no patterns suggesting unauthorized access to sensitive local files or transmission of data to external domains.
Audit Metadata