skills/thilinatlm/agent-skills/droid/Gen Agent Trust Hub

droid

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill automatically manages its own dependencies at runtime using a standard package manager.
  • Evidence: The scripts/droid-cli/droid and scripts/droid-cli/droid.ps1 wrapper scripts execute bun install if the node_modules directory is not found locally, fetching packages defined in package.json.
  • [COMMAND_EXECUTION]: The skill relies on executing external binaries to interact with Android hardware.
  • Evidence: The scripts/droid-cli/src/lib/adb.ts library uses Bun.spawn with an array of arguments to execute the adb (Android Debug Bridge) binary. This method is secure against host-side shell injection as it bypasses shell interpretation of arguments.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted UI data from third-party Android applications.
  • Ingestion points: scripts/droid-cli/src/lib/ui-hierarchy.ts captures the Android device UI hierarchy via the uiautomator dump command.
  • Boundary markers: UI elements are returned as structured data in a JSON array (elements), but the text content of these elements is not wrapped in markers to warn the agent about potentially malicious instructions embedded in the UI.
  • Capability inventory: The skill provides extensive capabilities to control the device, including tapping, typing, launching apps, and sending arbitrary key events.
  • Sanitization: No sanitization or filtering is performed on the text retrieved from the device UI before it is provided to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 07:11 AM