The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands for static and dynamic analysis of macOS binaries. Tools include system-native utilities such as file, codesign, spctl, otool, nm, and log, as well as third-party security tools like jtool2, lldb, and Frida.
[PROMPT_INJECTION]: The skill processes untrusted external data (Mach-O binaries, application bundles, and library dependencies) which can contain malicious metadata or embedded strings designed to influence agent behavior.
Ingestion points: Binary files (Mach-O, dylib, framework), app bundles, and LaunchAgent/Daemon configuration files specified as target or ./app in SKILL.md and references/macho-triage.md.
Boundary markers: No specific boundary markers or instructions to ignore embedded content are present when reading data from these sources.
Capability inventory: The skill possesses extensive command execution capabilities via shell utilities and debugging tools like lldb and Frida across all analyzed files.
Sanitization: No explicit sanitization or validation of input data or tool output is documented.

macos-reverse