cutedsl

Community

Write, debug, and optimize GPU kernels with CuTe.

AuthorShekswess
Version1.0.0
Installs0

System Documentation

What problem does it solve?

NVIDIA CuTe DSL enables writing CUTLASS-grade GPU kernels in Python. It compiles through AST rewrite + tracing → MLIR → ptxas, delivering the same PTX/SASS quality as C++ CUTLASS with dramatically faster compilation.

Core Features & Use Cases

  • Guides users on how to approach CuTe DSL tasks — the decision sequences, workflow, and judgment calls, not just the API surface.
  • Points to architecture-specific code patterns and API details in the official references, including references/api_reference.md and references/kernel_patterns.md.
  • Helps with porting C++ CUTLASS kernels to Python CuTe DSL, debugging, and optimizing kernels.

Quick Start

Provide a CuTe DSL kernel outline or review a CuTe DSL snippet and suggest improvements.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: cutedsl
Download link: https://github.com/Shekswess/gpu-kernel-skills/archive/main.zip#cutedsl

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.