z-rust-gtk4
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for building user interfaces that ingest untrusted data through various components. This constitutes a potential attack surface if the application logic does not properly sanitize these inputs before use in sensitive operations.
- Ingestion points: Signal handlers for
gtk::Entryandadw::EntryRow(specificallyconnect_changed) documented inSKILL.mdandreferences/adwaita-widgets.md. - Boundary markers: No delimiters or explicit instructions for handling untrusted content are included in the provided UI templates.
- Capability inventory: The skill demonstrates network operations (e.g.,
fetch_items), model state mutation, and system tray management via theksnicrate. - Sanitization: No input validation or sanitization routines are implemented in the boilerplate code.
Audit Metadata