skills.rest
dotruru
dotruruCOMMUNITY
Explore all skills by this author →
GitHub

cutlass-skill

Automates development, debugging, and optimization of CUTLASS and CuTeDSL GPU kernels using local sources.

Updated Apr 14, 2026
Search Tags:#debugging#cuda#cutlass#cute-dsl#kernel-optimization#gpu-kernels#header-reference
One-click install
npx skills add https://github.com/dotruru/claudemd --skill cutlass-skill-dotruru
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cutlass-skill
Source: https://github.com/dotruru/claudemd/tree/main/skills/cutlass
Command: npx skills add https://github.com/dotruru/claudemd --skill cutlass-skill-dotruru

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps developers write, debug, and optimize GPU kernels using CUTLASS and CuTeDSL by providing access to local source code, example workflows, and header references, reducing the time to understand and modify high-performance CUDA code.

Core Features & Use Cases

  • Local source navigation: Access and understand CUTLASS and CuTeDSL repository layouts and header files.
  • Compile & debug workflows: Build and run selective CUTLASS examples across CuTeDSL stacks to diagnose template errors and performance issues.
  • Header references & templates: Use header references and DSL bindings to explore kernel configurations and epilogue options.
  • Learning & code exploration: Explore CuTe tutorials and example workflows to learn CuTeDSL and CuTe integration patterns.

Quick Start

Load the CUTLASS and CuTeDSL sources from the local repos and begin by compiling a CuTeDSL example to observe the workflow.

Frequently Asked Questions about cutlass-skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug template errors when compiling CUTLASS GPU kernels?▼

To debug template errors in CUTLASS GPU kernels, you can build and run selective examples across CuTeDSL stacks to diagnose compilation issues using local source navigation and header references.

What is CuTeDSL and how does it integrate with CUTLASS for kernel optimization?▼

CuTeDSL provides bindings and templates that integrate with CUTLASS to explore kernel configurations and epilogue options, enabling developers to model realistic workflows for high-performance CUDA kernel development.

Do I need local CUTLASS and CuTeDSL repositories to optimize GPU kernels?▼

Yes, you need access to local repos containing CUTLASS and CuTe sources, CuTeDSL bindings, and corresponding headers to compile examples and model realistic kernel optimization workflows.

How do I navigate CUTLASS header references to configure GPU kernel epilogues?▼

You can navigate CUTLASS header references by accessing the local repository layout and using DSL bindings to explore available kernel configurations and epilogue options for your GPU kernels.

What's the best way to learn CuTe integration patterns for CUDA kernel development?▼

The best way to learn CuTe integration patterns is by exploring CuTe tutorials and example workflows within the local sources, then compiling CuTeDSL examples to observe the development workflow.

Why does my CUTLASS kernel compilation fail with complex template errors?▼

Complex template errors during CUTLASS kernel compilation often stem from misconfigured kernel parameters or epilogue options, which can be diagnosed by building selective examples across the CuTeDSL stack.

Related Skills

Explore More Software Engineering Skills→
slowlyCslowlyC
154

cutlass-skill

Develop, debug, and optimize CUTLASS and CuTeDSL GPU kernels for NVIDIA architectures.

Community
Advanced
LancerLabLancerLab

croq-dsl-cute-dsl

Tune and validate Python JIT GPU kernels using CuTe DSL.

Official
Intermediate
spacecat2002spacecat2002

add-sgl-kernel

Add AOT CUDA or C++ kernels to sgl-kernel with CMake and pytest.

Community
Advanced
jstzwjjstzwj
4

cutlass

Execute architecture-tuned GPU matrix multiplication kernels via NVIDIA CUTLASS.

Community
Advanced