android-ui-verification-v2
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on shell commands to interact with Android devices through the ADB (Android Debug Bridge) utility. Evidence includes calls to
adb shell,adb pull, andadb logcatwithin thescripts/verify_ui.shscript and the workflow described inSKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted UI data from mobile applications, which can serve as a vector for indirect prompt injection.
- Ingestion points: The script
scripts/verify_ui.shpulls a UI XML dump (/sdcard/view.xml) from the Android device to the local system. - Boundary markers: None; the instructions do not provide delimiters or specific warnings to treat the extracted XML content as potentially untrusted data.
- Capability inventory: The skill is capable of executing shell commands (
adb) and performing local file operations. - Sanitization: None; the skill does not perform any validation or sanitization of the UI hierarchy content before the agent is expected to parse or interact with it.
Audit Metadata