adb-navigation-base

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The Python scripts adb-tap.py, adb-swipe.py, and adb-wait-for.py execute adb shell commands via the subprocess module to perform UI interactions and screen captures. Coordinates and direction parameters are validated as integers or predefined choices to prevent command injection.
  • [INDIRECT_PROMPT_INJECTION]: The skill integrates OCR capabilities to monitor screen content, which represents a potential indirect prompt injection vector. Ingestion points: screenshots captured and processed by pytesseract in adb-wait-for.py and adb-tap.py. Boundary markers: No delimiters are used to separate screen text from system instructions. Capability inventory: The skill can execute various UI and application management commands on the device. Sanitization: Screen text is processed for verification but not sanitized against malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 12:32 PM
Security Audit — agent-trust-hub — adb-navigation-base