skills.rest
moonshineTP
moonshineTPCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

karpathy-guideline

Provide behavioral coding guidelines for LLM-assisted development.

Updated Apr 1, 2026
Search Tags:#code quality#code review#best practices#refactoring#LLM coding
One-click install
npx skills add https://github.com/moonshineTP/ATTT-20252 --skill karpathy-guideline
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: karpathy-guideline
Source: https://github.com/moonshineTP/ATTT-20252/tree/main/.gemini/skills/karpathy-guideline
Command: npx skills add https://github.com/moonshineTP/ATTT-20252 --skill karpathy-guideline

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill addresses common mistakes in coding with Large Language Models (LLMs), promoting cleaner, more efficient code.

Core Features & Use Cases

  • Code Best Practices: Offers guidelines for avoiding overcomplication, making surgical changes, and defining verifiable success criteria.
  • Surgical Changes: Emphasizes touching only necessary code and not improving adjacent code without cause.
  • Goal-Driven Execution: Encourages setting clear success criteria and verifying each step until the goal is achieved.
  • Use Case: Ideal for coders looking to review, refactor, or write code, aiming to maintain simplicity and avoid pitfalls common in LLM development.

Quick Start

Use the karpathy-guideline skill to follow Andrej Karpathy's guidelines when coding with LLMs to minimize mistakes.

Frequently Asked Questions about karpathy-guideline

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

FAQPage Schema
What are the best practices for coding with LLMs to avoid common mistakes?▼

Coding with LLMs requires behavioral guidelines to minimize errors, adopting a minimalist approach, making surgical changes, and defining verifiable success criteria to improve overall code quality.

How do I ensure LLM coding generates clean and efficient code?▼

To ensure clean and efficient LLM coding, apply goal-driven execution by setting clear success criteria and verifying each step until the goal is achieved, while actively avoiding overcomplication.

How do I perform surgical changes when refactoring code with an LLM?▼

Performing surgical changes when refactoring code with an LLM means touching only the necessary code without improving adjacent code without cause, maintaining simplicity and preventing overcomplication.

Does goal-driven execution improve LLM code review outcomes?▼

Goal-driven execution improves LLM code review outcomes by encouraging the definition of clear success criteria and verifying each development step until the specific goal is achieved.

When should I not use an LLM to improve adjacent code during refactoring?▼

You should not use an LLM to improve adjacent code during refactoring without a specific cause, as surgical changes emphasize touching only necessary code to prevent overcomplication and maintain simplicity.

Related Skills

Explore More Software Engineering Skills→
sQVesQVe

write-plan

Convert a brainstorm design summary into an immutable epic with a first task.

Community
Advanced
xdgxdg
20

refactor

Apply surgical refactoring to improve code quality without changing behavior.

Community
Intermediate
ProgrammerAnthonyProgrammerAnthony
138

frontend-code-review

Analyze React and TypeScript code for structural quality and performance bottlenecks.

Community
Intermediate
dushyantkhosladushyantkhosla

karpathy-guidelines

Apply behavioral coding guidelines for simplicity and surgical changes.

Community
Intermediate