The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell commands (e.g., ps, df, uptime, system_profiler, powershell) to collect system health metrics across macOS, Windows, and Linux. This execution is limited to information gathering through the start_process tool and aligns with the skill's stated purpose.- [INDIRECT_PROMPT_INJECTION]: The skill analyzes command output from the local system, which serves as an ingestion point for untrusted data (e.g., malicious process names or filenames). The risk is addressed through the skill's focus on quantitative metrics and bounded output processing.
Ingestion points: System diagnostic output gathered from scripts in references/linux.md, references/macos.md, and references/windows.md.
Boundary markers: None explicitly defined to separate command output from the prompt context.
Capability inventory: Shell process control via start_process and interact_with_process as defined in the skill's tool configuration.
Sanitization: Employs truncation commands like head and Select-Object -First to limit the volume of data ingested into the agent context.

computer-health-check