android-cli
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: Fetches and executes installation scripts from official Google domains (
dl.google.com). The skill provides commands to pipe shell scripts directly to bash for Linux and macOS environments, and to download and execute a batch file for Windows installations. - [EXTERNAL_DOWNLOADS]: Downloads the
androidCLI tool, SDK components, and potentially additional plugin skills via theandroid sdk installandandroid skills addcommands from established repositories. - [COMMAND_EXECUTION]: Executes local shell commands to interact with Android devices via
adb shell inputand manages the development environment through variousandroidCLI subcommands such ascreate,emulator, andlayout. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill ingests untrusted data in the form of XML-based 'journey' test files (
references/journeys.md) which contain natural language instructions for the agent to follow. - Boundary markers: Natural language actions are delineated by
<action>XML elements. - Capability inventory: The skill possesses capabilities to interact with hardware devices (
adb), write files (screenshots), and manage software packages (android sdk). - Sanitization: No specific filtering or sanitization of the natural language content within the XML action tags is documented.
Audit Metadata