android-ui-journey-testing
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the XML journey specifications it processes. Maliciously crafted
<action>tags could contain instructions designed to override agent behavior or bypass constraints. - Ingestion points: XML test specifications (e.g.,
<journey>,<actions>) described in Step 1. - Boundary markers: None specified; instructions do not require delimiters or specific 'ignore' warnings for the XML content.
- Capability inventory: Execution of shell commands via ADB (
adb shell input tap,adb shell input text, etc.). - Sanitization: No sanitization or validation of the XML action strings is described before processing.
- [COMMAND_EXECUTION]: The skill constructs and executes shell commands (
adb shell input ...) using parameters extracted from the XML actions. While this is the primary functionality of the tool, executing shell commands based on untrusted external data carries a risk of command injection or unintended device manipulation if the inputs are not properly sanitized.
Audit Metadata