The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection by ingesting untrusted data from mobile applications during the testing process.\n
Ingestion points: Decompiled source code from jadx_decompile, system log output from adb_command, and raw files retrieved from the mobile device's data directory (references/tools.md, references/workflows.md).\n
Boundary markers: The instructions lack explicit delimiters or warnings to treat ingested application data as untrusted.\n
Capability inventory: The skill provides capabilities for shell command execution via adb_command and dynamic code injection via frida_hook (references/tools.md).\n
Sanitization: No sanitization or validation of external content is documented before processing.\n- [COMMAND_EXECUTION]: The skill defines an adb_command function that allows the execution of arbitrary shell commands on connected mobile devices, which could be exploited to perform unintended operations on the target system (references/tools.md).\n- [DYNAMIC_EXECUTION]: The skill utilizes frida_hook to execute JavaScript instrumentation scripts at runtime to intercept and modify application behavior, representing a dynamic execution capability (references/tools.md).

pentest-mobile-app