jamboree-android-security-testing
Fail
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Clones an entire testing framework from a non-trusted third-party repository at
https://github.com/hero-mike/Android-Mobile-Security-Sandbox-Testing.git(SKILL.md). - [REMOTE_CODE_EXECUTION]: Instructs the agent to execute shell scripts directly from the cloned repository, such as
./orchestration/validators/check-environment.shand./orchestration/deployers/install-all.sh(SKILL.md). - [COMMAND_EXECUTION]: Executes potentially dangerous system commands including
adb root,adb remount, andadb shell settings put global http_proxyto modify device behavior and gain privileged access (SKILL.md). - [DYNAMIC_EXECUTION]: Employs Frida and Objection to dynamically inject and execute JavaScript code into running processes for SSL pinning bypass and method interception (SKILL.md).
- [PERSISTENCE]: Recommends modifications to user shell profiles (e.g.,
~/.bashrc) or.envfiles to store environment variables and configuration paths (SKILL.md). - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted output from target applications and Android system tools.
- Ingestion points: Data is ingested through
frida-ps,adb shell, andobjectionexploration commands (SKILL.md). - Boundary markers: No boundary markers or instructions to ignore embedded content are used when processing external data.
- Capability inventory: The skill has the ability to execute shell commands via
adb, write files to the system, and redirect network traffic (SKILL.md). - Sanitization: No sanitization or validation of data retrieved from the Android environment is performed.
Recommendations
- AI detected serious security threats
Audit Metadata