Open Skills Logo
Open Skills

get-available-resources

This skill should be used at the start of any computationally intensive scientific task to detect and report available system resources (CPU cores, GPUs, memory, disk space). It creates a JSON file with resource information and strategic recommendations that inform computational approach decisions such as whether to use parallel processing (joblib, multiprocessing), out-of-core computing (Dask, Zarr), GPU acceleration (PyTorch, JAX), or memory-efficient strategies. Use this skill before running analyses, training models, processing large datasets, or any task where resource constraints matter.

Author

@K-Dense-AI

Category

Development Tools

Install

Hot:4

Download and extract to your skills directory

Copy command and send to AI Agent for auto-install:

Download and install this skill https://openskills.cc/api/download?slug=k-dense-ai-skills-get-available-resources&locale=en&source=copy

Get Available Resources - System Compute Resource Detection

Skill Overview


Get Available Resources is a resource detection skill designed specifically for scientific computing. Before starting compute-intensive tasks, it automatically checks CPU, GPU, memory, and disk space, and generates strategic recommendations to help you choose an appropriate computing approach (e.g., Dask, PyTorch, joblib, etc.).

Use Cases

1. Before Large-Scale Data Analysis


When working with GB-scale datasets, use this skill to detect available memory and disk space. It automatically determines whether you should use Pandas (in-memory loading) or Dask (chunked processing), avoiding program crashes caused by insufficient memory.

2. Before Machine Learning Training


Before starting model training, detect GPU availability (supports NVIDIA CUDA, AMD ROCm, Apple Silicon Metal) and the backend. It recommends suitable deep learning frameworks (PyTorch, TensorFlow, JAX) and device configurations.

3. Parallel Processing Optimization


Before running large-scale parallel tasks, it automatically detects the number of CPU cores and recommends the optimal worker count. It provides parallelization strategies for joblib, multiprocessing, and Dask to improve performance.

Core Features

1. Comprehensive Hardware Detection


Automatically detects system hardware information, including:
  • CPU (physical/logical core counts, architecture),

  • GPU (NVIDIA, AMD, Apple Silicon, along with their VRAM and driver versions),

  • Memory (total and available capacity, usage rate, swap space),

  • Disk (total and available space),

    • as well as operating system and Python version details.

    2. Intelligent Resource Recommendations


    Generates context-aware strategic recommendations based on the detection results, including:
  • Parallel processing strategies (high/medium/low parallelism levels and recommended worker counts),

  • Memory strategies (memory-limited / moderate / sufficient, and recommended libraries such as Dask/Zarr),

  • GPU acceleration recommendations (available backends and corresponding libraries),

  • Large-data processing strategies.

    • 3. Cross-Platform GPU Support


    Fully supports the three major GPU platforms:
  • Detects NVIDIA GPUs via nvidia-smi and reports VRAM and compute capability,

  • Detects AMD GPUs via rocm-smi,

  • Uses system detection to identify Apple Silicon M1/M2/M3/M4 chips and their unified memory support through Metal.

    • Frequently Asked Questions

    How do I detect which GPUs are available on my system?


    Run the scripts/detect_resources.py script provided by the skill. It automatically detects NVIDIA, AMD, and Apple Silicon GPUs and reports GPU models, VRAM size, driver versions, and available backends (CUDA, ROCm, Metal) in the generated .claude_resources.json file.

    When should I use Dask instead of Pandas?


    When the dataset size exceeds 50% of the available memory, the skill recommends using Dask for out-of-core processing. For example, if available memory is 8GB and the dataset is 10GB, use Dask for chunked loading instead of loading everything at once with Pandas.

    Can Apple Silicon Macs use GPU acceleration?


    Yes. The skill detects Apple M1/M2/M3/M4 chips and their Metal support, and recommends using the PyTorch MPS backend, TensorFlow-Metal, or JAX-Metal for GPU acceleration to fully leverage the unified memory architecture.