The Agent Skills Directory

[COMMAND_EXECUTION]: The security_suite.py script executes various system profiling tools, including file, otool, ldd, strings, ps, and lsof, to collect static and dynamic information about a target binary. This is standard and expected behavior for a security analysis utility.
[COMMAND_EXECUTION]: The skill uses subprocess.Popen to execute a user-specified binary within a controlled environment to observe its runtime behavior, such as file system modifications and network endpoint usage. The script employs environment sanitization for this dynamic trace.
[SAFE]: The prompt_redteam.py script performs offline scans of repository files using regular expressions to detect potential security regressions or unsafe patterns. It operates exclusively on local files without network interaction or arbitrary code execution.
[SAFE]: The implementation demonstrates security best practices, such as utilizing shlex.quote to sanitize arguments for shell-based command lookups and implementing a sandboxed context for dynamic binary execution.

security-suite