Get Available Resources

Before you spin up that neural network training run or load a 50GB dataset, this does the smart thing: checks what you actually have to work with. It scans for CPU cores, GPU availability (NVIDIA, AMD, or Apple Silicon), memory, and disk space, then writes a JSON file with concrete recommendations like "use 6 workers for joblib" or "this dataset needs Dask because you only have 8GB free." The recommendations are genuinely helpful, telling you whether to go parallel, use out-of-core processing, or tap into GPU acceleration based on what it finds. It's the kind of sanity check that prevents you from writing code that immediately falls over because you assumed more resources than you have.