ako4all

Official

Automates end-to-end GPU kernel optimization.

AuthorTongmingLAIC
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Iterative, automated optimization of GPU kernels (CUDA, Triton, TileLang, or C/C++) to maximize runtime speedups in a drop-in workflow.

Core Features & Use Cases

  • Establishes a workspace and runs a repeatable optimize-benchmark-improve loop (profiling with ncu, correctness checks, and iterative edits).
  • Supports multiple languages/backends and a plug-in bench (KernelBench) with built-in evaluation and logging to ITERATIONS.md.
  • Commits progress as a traceable artifact via git, enabling resumable optimization across sessions.

Quick Start

Open Claude Code in a directory containing your kernel and instruct it to start an optimization loop.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 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: ako4all
Download link: https://github.com/TongmingLAIC/AKO4ALL/archive/main.zip#ako4all

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 510,000+ vetted skills library on demand.