karpathy-guidelines

Provide behavioral guidelines for LLM coding to avoid overcomplication and hidden assumptions.

12|1|Updated Mar 15, 2026

Search Tags:#code quality #code review #refactoring #behavioral guidelines #LLM coding

Ready to run

npx skills add https://github.com/jorgeasaurus/agent-skills --skill karpathy-guidelines-jorgeasaurus

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: karpathy-guidelines
Source: https://github.com/jorgeasaurus/agent-skills/tree/main/karpathy-guidelines
Command: npx skills add https://github.com/jorgeasaurus/agent-skills --skill karpathy-guidelines-jorgeasaurus

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The karpathy-guidelines Skill helps in reducing common LLM coding mistakes by providing behavioral guidelines.

Core Features & Use Cases

Behavioral Guidelines: Offers guidelines to avoid overcomplication, hidden confusion, and undefined success criteria.
Surgical Changes: Encourages making changes only where necessary and cleaning up only your own mess.
Goal-Driven Execution: Emphasizes defining verifiable success criteria and looping until verified.

Quick Start

Run the karpathy-guidelines Skill to review your code and follow the guidelines to improve its quality.

Frequently Asked Questions about karpathy-guidelines

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

FAQPage Schema

How do I prevent overcomplication when using LLM coding for refactoring?▼

To prevent overcomplication during LLM coding refactoring, apply behavioral guidelines that encourage making surgical changes only where necessary and cleaning up only your own mess. This approach ensures targeted modifications without unnecessary code bloat.

What are common LLM coding pitfalls during code review?▼

Common LLM coding pitfalls during code review include hidden assumptions, undefined success criteria, and overcomplication. You can avoid these by applying behavioral guidelines that enforce clear goals and surgical changes to improve overall code quality.

How do I define verifiable success criteria for LLM generated code?▼

Defining verifiable success criteria for LLM generated code requires goal-driven execution where you establish clear metrics upfront. You then loop the LLM coding process until those specific criteria are verified as met.

Can I use behavioral guidelines to fix hidden assumptions in LLM code?▼

Yes, you can use behavioral guidelines to fix hidden assumptions in LLM code. The guidelines specifically target hidden confusion by enforcing explicit definition of success criteria and promoting surgical changes during code review and refactoring.

Does applying surgical changes improve LLM code quality?▼

Applying surgical changes directly improves LLM code quality by ensuring modifications are made only where necessary. This behavioral guideline prevents overcomplication and confines cleanup to your own mess, resulting in cleaner refactoring outputs.

When should I use behavioral guidelines for LLM coding?▼

You should use behavioral guidelines for LLM coding whenever you need to avoid common pitfalls like overcomplication and hidden assumptions. They are specifically applicable during code reviews and refactoring to enforce clear success criteria and surgical changes.