What is kernel-skills CLAUDE.md?

A set of repository instructions for kernel-skills, a public library of guidance files for writing compute kernels. Compute kernels are low-level programs designed to perform numerical work efficiently, including on GPUs.

Use it when working on guidance for CUDA or Triton GPU code, quantization-related kernels, or other high-performance numerical code.

It tells coding agents what the repository is for, what belongs in it, and how its guidance should improve the correctness and performance awareness of generated kernel code.

How tensormux/kernel-skills configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to tensormux/kernel-skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/tensormux/kernel-skills/master/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/tensormux/kernel-skills

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/tensormux/kernel-skills/claude-md/github.svg)](https://agentmods.dev/instructions/tensormux/kernel-skills/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/tensormux/kernel-skills/claude-md"><img src="https://agentmods.dev/badge/instructions/tensormux/kernel-skills/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/tensormux/kernel-skills/claude-md"><img src="https://agentmods.dev/badge/instructions/tensormux/kernel-skills/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 4,586 This file is loaded in full into every session.

When invoked 4,586 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.04586	$0.04586
Opus 5.5	$0.01834	$0.01834
Sonnet 5.5	$0.00917	$0.00917
Haiku 4.5	$0.00459	$0.00459

Measured 2mo ago against content hash 0053db5c18eb, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

kernel-skills CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 718 lines

How it starts

The opening of the file, as written. The whole thing — 718 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

kernel-skills

Repository mission

kernel-skills is an open source repository of extremely high quality skill files for AI coding agents.

The purpose of this repository is to help agents write better compute kernels across CUDA, Triton, quantization-oriented kernels, and core optimization patterns. This repository is not a runtime, framework, package manager, or MCP server in its first iteration. It is a curated and disciplined library of SKILL.md files that agents and humans can directly use.

The repository must be public, contribution-friendly, easy to browse, and immediately useful without any tooling beyond reading and copying Markdown files.

The central principle is simple:

Every skill in this repository must materially improve the quality, correctness, and performance-awareness of kernel code produced by an AI agent.

What this repository is

This repository is:

a library of structured SKILL.md files
focused on kernel engineering for AI and high performance numerics
designed for use with coding agents like Claude Code, ChatGPT, Cursor, Gemini CLI, Codex, and similar tools
optimized for practical reuse and public OSS contribution
strict about quality, technical accuracy, and clarity

This repository is not:

a kernel runtime
a benchmark harness platform
an MCP integration project
a code execution framework
a dumping ground for vague prompts
a tutorial-only repo
a place for low-quality or repetitive skills

Product scope for v1

Focus only on skills.

Do not build packaging systems, CLI tools, MCP servers, benchmark runners, or automation infrastructure in the first iteration unless absolutely required for basic repo hygiene.

The first version should prove one thing:

well-authored skill files can help AI agents produce significantly better kernel code

Core outcome we want

A user should be able to open any skill in this repository, paste it into their agent workflow, and get a much better result than they would get from a generic prompt.

Read the full file on GitHub · 718 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2mo ago First seen · 718 lines · 4,586 tokens per session scan A 0053db5c18eb

Subscribe to this mod's changes

kernel-skills CLAUDE.md is an instructions file published in the GitHub repository tensormux/kernel-skills (89 stars, last pushed 4mo ago), licensed MIT. It adds 4,586 tokens to every session, about $0.0183 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

ai-coding cuda gpu-kernels high-performance-computing kernel-optimization llm-agents

kernel-skills CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
kernel-skills CLAUDE.mdtensormux/kernel-skills	Instructions file	89	4mo ago	MIT	A	4,586	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

kernel-skills: Instructions file for Claude Code