The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The scripts health_check.py, api_bench.py, and monitor.py automatically attempt to install the psutil package from the Python Package Index (PyPI) if it is not found on the system.
[COMMAND_EXECUTION]: The skill uses subprocess.check_call to execute shell commands for package installation.
[DATA_EXPOSURE]: The skill accesses detailed system information, including:
Full process lists and their associated memory/CPU usage via psutil.process_iter.
Local disk usage statistics for the C:/ drive.
Network interface statistics and MTU settings.
Local file system paths which are hardcoded in the SKILL.md file, revealing the author's local directory structure and username.
[INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection:
Ingestion points: health_check.py and monitor.py read process names directly from the operating system.
Boundary markers: No delimiters or safety instructions are used when interpolating these names into reports.
Capability inventory: The skill possesses the ability to write files to disk (monitor_log.json) and execute subprocesses (for pip installation).
Sanitization: Process names and system metrics are not sanitized or validated before being included in the diagnostic summaries shown to the agent and the user.

claude-monitor