android-pentest-environment-setup

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches several third-party security frameworks and tools from well-known hosting services.
  • Downloads Magisk, Frida, LSPosed, ZygiskNext, and Shamiko from GitHub repositories.
  • Clones the rootAVD tool from GitLab for emulator patching.
  • [REMOTE_CODE_EXECUTION]: The installation process involves executing scripts and binaries downloaded from external sources.
  • Executes rootAVD.sh after cloning from GitLab to modify the emulator's ramdisk.
  • Downloads and runs the frida-server binary on the Android device via ADB.
  • [COMMAND_EXECUTION]: The skill performs various system-level command operations to configure the environment.
  • Appends environment variable exports (ANDROID_HOME, PATH) to shell profile files (~/.bash_profile, ~/.zshrc).
  • Uses avdmanager and sdkmanager to create and manage virtual devices and system images.
  • [PRIVILEGE_ESCALATION]: The skill instructions require high-level permissions to achieve a rooted state.
  • Utilizes adb root and adb shell su to gain administrative access to the emulator.
  • Modifies the system partition (adb remount) to install custom CA certificates and Magisk modules.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes a high-privilege environment designed to process untrusted application data and dynamic instrumentation scripts.
  • Ingestion points: The AVD environment is designed to ingest and execute external APKs and Frida instrumentation scripts (SKILL.md).
  • Boundary markers: No explicit delimiters or warnings for the agent to ignore instructions embedded in the tested applications are provided.
  • Capability inventory: The environment provides full shell access, root privileges, and network manipulation capabilities across multiple scripts.
  • Sanitization: No sanitization or validation mechanisms are described for the external components being integrated into the environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 05:07 PM