visual-wayfinder
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust security model for interactive UI, ensuring that all rendered content is confined to a strictly allowlisted set of components without execution capabilities.
- [PROMPT_INJECTION]: Comprehensive instructions are provided to maintain boundaries between the agent and the decision tracker, explicitly forbidding the AI from generating logic that could circumvent safety filters or user confirmation.
- [DATA_EXFILTRATION]: No network operations or data harvesting patterns were identified. The architecture prevents the exposure of sensitive ticket data to the UI rendering prompt.
- [COMMAND_EXECUTION]: The skill explicitly excludes components capable of triggering commands or tool calls from the generated canvas, ensuring that only host-owned controls can initiate system actions.
- [EXTERNAL_DOWNLOADS]: All dependencies are referenced as standard platform skills, and the interactive demo is self-contained with no external script loading.
Audit Metadata