The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute system monitoring commands including ps, awk, sysctl, uptime, and vm_stat. These are standard utilities for performance analysis.
[COMMAND_EXECUTION]: Provides instructions for the agent to use kill and killall commands to terminate processes. The skill includes safety constraints requiring explicit user confirmation before execution and forbids terminating critical system processes.
[DATA_EXFILTRATION]: The skill collects a full list of running processes (ps aux). This may expose sensitive information if applications are running with credentials or secrets in their command-line arguments. No network transmission or external exfiltration patterns were detected.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data originating from the operating system's process list.
Ingestion points: System command output from ps aux and process groupings in SKILL.md.
Boundary markers: None identified; output is formatted into markdown tables for the user.
Capability inventory: The agent has the ability to execute kill and killall commands via the Bash tool.
Sanitization: No explicit sanitization or filtering of process names or command-line arguments is performed before the data is processed by the agent's logic.

system-doctor