The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/detect_resources.py executes several system diagnostic utilities to gather hardware specifications.
Evidence: The script uses subprocess.run() to invoke nvidia-smi (for NVIDIA GPU details), rocm-smi (for AMD GPU details), sysctl (for macOS processor strings), and system_profiler (for Apple Silicon GPU info).
Context: These commands are used to identify available hardware backends. The script implements these calls using static argument lists without shell interpolation or user-input concatenation, which is a safe practice for this type of system discovery.
[EXTERNAL_DOWNLOADS]: The skill documentation specifies a dependency on an external Python library.
Evidence: The SKILL.md file instructs users to install psutil via uv pip install psutil.
Context: psutil is a well-established and trusted industry-standard library for cross-platform system monitoring and resource management.

get-available-resources