ios-pentesting-tricks

Pass

Audited by Gen Agent Trust Hub on Nov 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a variety of shell commands for analyzing iOS binaries and application environments using tools such as otool, plutil, ideviceinstaller, ios-deploy, sqlite3, and grep.
  • [DYNAMIC_EXECUTION]: The skill includes extensive recipes for Frida and Objection to perform runtime manipulation, including Objective-C method hooking, swizzling, and modifying return values in live processes.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources that could potentially contain instructions intended to influence the agent's behavior.
  • Ingestion points: Fetches AASA files from remote servers (e.g., target.com) and reads application Info.plist and sandbox files.
  • Boundary markers: No specific delimiters or instructions to ignore embedded content are used when processing external data.
  • Capability inventory: Subprocess execution of various CLI tools, runtime code injection via Frida, and filesystem access.
  • Sanitization: The instructions do not specify any validation or sanitization for the retrieved AASA or plist data.
  • [EXTERNAL_DOWNLOADS]: The skill references the use of external tools and scripts such as frida-ios-dump, bagbak, and dump.py for tasks like binary decryption and extraction.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 16, 2026, 12:04 PM