design-ui-window
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses structured, task-specific instructions to guide UI redesign and does not contain any patterns attempting to bypass agent safety or extract system prompts.\n- [DATA_EXFILTRATION]: No indicators of unauthorized data access or external transmission were found. The skill references local project-specific UI files and theme resources, which is consistent with its stated design purpose.\n- [REMOTE_CODE_EXECUTION]: The skill does not download external dependencies or execute remote scripts. It focuses entirely on XAML-based UI modifications.\n- [SAFE]: The skill adheres to its defined functionality without requesting excessive permissions or performing suspicious operations. It utilizes vendor-specific project paths that align with the author's identity.
Audit Metadata