소개
The Computational Resource Profiler skill allows Claude to perform a comprehensive audit of local hardware environments before initiating resource-heavy operations. By identifying specific CPU core counts, GPU backends (including NVIDIA CUDA, AMD ROCm, and Apple Silicon Metal), and memory availability, it generates a structured JSON report with actionable recommendations. This ensures that subsequent tasks—such as training machine learning models or processing large datasets—automatically utilize the most efficient libraries and parallelization strategies tailored to the specific machine, preventing memory errors and maximizing performance.
