How do I apply the recommendations provided by the skill?

The skill generates a '.claude_resources.json' file which contains specific strategy keys for parallel processing, memory, and GPUs that you can use to configure your Python scripts.

What GPU backends does this skill detect?

It detects NVIDIA CUDA via nvidia-smi, AMD ROCm via rocm-smi, and Apple Silicon (M1-M4) via Metal and unified memory reporting.

Can this skill help with OOM (Out of Memory) errors?

Yes, by detecting available RAM and comparing it to dataset size, it recommends out-of-core libraries like Dask or Zarr before you start processing.

Does this skill require any external Python packages?

Yes, it requires the 'psutil' package which can be installed via 'uv pip install psutil' or standard pip.

System Resource Detector & Advisor

Name: System Resource Detector & Advisor
Author: Weiwei-Mao

byWeiwei-Mao

0•

Data Science & ML

Detects hardware resources and provides strategic architectural recommendations for computationally intensive scientific tasks.

The get-available-resources skill identifies system capabilities including CPU cores, GPU backends (NVIDIA CUDA, AMD ROCm, Apple Metal), RAM, and disk space to generate actionable computing strategies. It helps developers and data scientists make informed decisions about using parallel processing (joblib, Dask), out-of-core computing (Zarr), and hardware acceleration (PyTorch, JAX). By running this skill at the start of a project, you can ensure your implementation is optimized for your specific hardware constraints, avoiding memory overflows and maximizing performance.

Key Features

01Comprehensive hardware detection including NVIDIA, AMD, and Apple Silicon GPUs

02Strategic recommendations for parallel processing and worker counts

03Generates portable JSON configuration for programmatic tool integration

04Cross-platform support for macOS, Linux, and Windows environments

050 GitHub stars

06Context-aware memory management strategies for large datasets

Use Cases

01Optimizing worker counts for joblib or multiprocessing tasks

02Selecting the appropriate GPU backend for deep learning model training

03Determining if a dataset requires Dask or Zarr for out-of-core processing

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add weiwei-mao/hydrology-skills get-available-resources

For use in Claude.ai and ChatGPT

Download Skill