The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents the use of standard system utilities such as lsof, ps, kill, du, and df. These tools are used for their intended purposes: identifying port usage, terminating unresponsive processes, and diagnosing disk space issues. While kill -9 and rm -rf are powerful commands, they are presented in a relevant context for resource management.
[DATA_EXFILTRATION]: No evidence of network exfiltration or unauthorized data movement. Network-related commands like lsof and ss are used exclusively for local port diagnostics.
[EXTERNAL_DOWNLOADS]: The documentation mentions installing common system utilities like htop, ncdu, and pstree through standard package managers (e.g., Homebrew). No direct remote script execution or unverified downloads were detected.
[PROMPT_INJECTION]: The skill was evaluated for indirect prompt injection surfaces. While it processes system data such as process names and file paths, there are no instructions that would cause the agent to deviate from its safety guidelines or execute hidden commands embedded in system metadata.
[CREDENTIALS_UNSAFE]: No hardcoded credentials, API keys, or instructions to expose sensitive configuration files were found.

port-process