The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill dynamically generates Frida JavaScript instrumentation code and executes it on the target device.
Evidence: scripts/agent.py uses f-strings to inject target class names into JavaScript templates before passing them to the Frida CLI.
[COMMAND_EXECUTION]: The Python scripts act as wrappers for Frida command-line utilities.
Evidence: Both scripts/agent.py and scripts/process.py use subprocess.run to invoke frida, frida-ps, and frida-trace.
[INDIRECT_PROMPT_INJECTION]: The skill consumes output from external applications (via Frida) which could potentially contain malicious instructions if the target app is designed to subvert the analysis tool.
Ingestion points: run_frida_script in process.py and _frida_cmd in agent.py capture standard output from the Frida process.
Boundary markers: Not present.
Capability inventory: The skill has the ability to write JSON reports to the local filesystem and execute shell commands.
Sanitization: The tool parses JSON output but does not explicitly sanitize strings extracted from the target application's runtime.

reverse-engineering-ios-app-with-frida