mobile-pentest

Pass

Audited by Gen Agent Trust Hub on May 26, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a library of CLI commands using adb, frida, objection, and various decompilers to audit mobile applications. These commands are standard for mobile security professionals and match the skill's stated intent.
  • [EXTERNAL_DOWNLOADS]: The documentation references industry-standard tools like Magisk, LSPosed, Jadx, and Apktool. No automated execution of remote scripts or unverified package installations were detected.
  • [PROMPT_INJECTION]: The skill describes processes for ingesting data from untrusted third-party mobile applications (e.g., parsing manifests and source code). This constitutes an indirect prompt injection surface.
  • Ingestion points: Reads and parses AndroidManifest.xml, decompiled Java source code, and binary strings from APK/IPA files.
  • Boundary markers: None provided in the command examples.
  • Capability inventory: Includes the ability to execute shell commands via adb, inject JavaScript into processes via Frida, and read/write local files.
  • Sanitization: No specific sanitization or validation of the extracted application data is described.
Audit Metadata
Risk Level
SAFE
Analyzed
May 26, 2026, 08:56 AM
Security Audit — agent-trust-hub — mobile-pentest