home/categories/framework-internals/invergent-ai-surogate-claude-skills-cuda-skill-md
framework-internalsdevelopment

cuda

Write or optimize CUDA kernels for the Surogate framework. Use when the user wants to create CUDA kernels, optimize existing kernels for specific GPU architectures (sm89/sm90/sm100/sm103/sm120/sm121), implement custom attention/normalization/activation layers, or debug kernel performance issues.

invergent-ai
maintainer
invergent-ai
Updated 3/10/2026
Stars
211
Forks
3
quick start

Installation and usage

Write or optimize CUDA kernels for the Surogate framework. Use when the user wants to create CUDA kernels, optimize existing kernels for specific GPU architectures (sm89/sm90/sm100/sm103/sm120/sm121), implement custom attention/normalization/activation layers, or debug kernel performance issues.

Installation
$ install --globalskills.sh
Usage

Once installed, you can use this skill by running the following command in your terminal:

skills use cuda