android_ui_verification
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes
adbcommands to interact with an Android emulator or device. This includes retrieving the UI hierarchy, capturing screenshots, and simulating user inputs like taps and swipes, which are standard for UI automation tools. - [PROMPT_INJECTION]: There is an attack surface for indirect prompt injection as the skill ingests external data from the device. Ingestion points: The skill pulls
view.xml(UI hierarchy) andlogcat(system logs) viascripts/verify_ui.sh. Boundary markers: None are present to wrap device-sourced data. Capability inventory: The skill can input text and trigger key events via ADB. Sanitization: No sanitization is performed on the data pulled from the device before it is processed by the agent.
Audit Metadata