gui-agent-mobile

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates Android GUI automation by invoking the gui_agent_skill module through CLI commands such as python -m gui_agent_skill.cli execute and python -m gui_agent_skill.cli tap. These commands are used to interact with mobile device interfaces and are central to the skill's legitimate functionality.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from the device screen to determine its next actions. Ingestion points: Screen captions and screenshot paths are returned by the CLI for analysis by the agent. Boundary markers: No specific delimiters or warnings to ignore instructions within the UI data are present in the skill instructions. Capability inventory: The skill can perform multi-step automation, UI clicks, and coordinate-based taps on a connected Android device. Sanitization: The skill does not describe any mechanisms for validating or sanitizing the content retrieved from the device screen before processing it.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 09:56 AM