cuopt-numerical-optimization-formulation
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely composed of Markdown instructions and conceptual guidelines. It contains no executable scripts, binaries, or configuration files that could trigger code execution.
- [SAFE]: External links found in the skill metadata and documentation point to official NVIDIA documentation and public GitHub repositories belonging to the NVIDIA organization. These are trusted vendor resources.
- [SAFE]: No mechanisms for data exfiltration, credential harvesting, or privilege escalation were identified.
- [SAFE]: Although the skill processes user-supplied problem descriptions, it lacks any execution tools or capabilities (such as subprocess calls or file writing), which mitigates the risk of indirect prompt injection.
- [SAFE]: The instructional language, including mandatory clarification steps, is designed to ensure accuracy and user confirmation rather than to bypass agent safety filters or guidelines.
Audit Metadata