karpathy-guidelines

Apply explicit thinking and minimal changes to coding tasks.

1|Updated Feb 13, 2026
One-click install
npx skills add https://github.com/paialex/ai-skills-bank --skill karpathy-guidelines-paialex
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: karpathy-guidelines
Source: https://github.com/paialex/ai-skills-bank/tree/main/skills/karpathy-guidelines
Command: npx skills add https://github.com/paialex/ai-skills-bank --skill karpathy-guidelines-paialex

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This set of behavioral guidelines helps reduce common LLM coding mistakes by enforcing explicit thinking, minimal changes, and verifiable success criteria during coding, reviewing, and refactoring.

Core Features & Use Cases

  • Provides explicit thought processes, surface tradeoffs, and plan-driven execution to guide developers.
  • Helps teams enforce minimal, surgical changes and maintain codebase integrity during edits.
  • Use case: when writing or reviewing code, apply these guidelines to surface assumptions and define success criteria before making changes.

Quick Start

Follow Karpathy Guidelines in your next coding task by explicitly stating assumptions, choosing a minimal approach, and outlining verifiable success criteria.

Frequently Asked Questions about karpathy-guidelines

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

FAQPage Schema
How do I reduce unnecessary complexity when using LLMs for coding tasks?

To reduce unnecessary complexity in LLM coding, apply explicit thinking processes and enforce minimal, surgical changes. Surface assumptions and define verifiable success criteria before writing or refactoring code to maintain codebase integrity.

What is the best way to plan surgical code changes during a refactoring workflow?

The best way to plan surgical code changes is to state assumptions explicitly, choose a minimal approach, and outline verifiable success criteria. This plan-driven execution guides developers through minimal, measurable edits during refactoring.

How do I define verifiable success criteria before reviewing code?

Defining verifiable success criteria involves explicitly stating assumptions and planning minimal edits before reviewing code. This surfaces tradeoffs and ensures LLM coding tasks produce measurable outcomes and verifiable results.

Why does my LLM generate overly complex code edits during refactoring?

LLMs generate complex code edits when lacking explicit thinking and plan-driven execution. Enforcing surgical changes and surfacing tradeoffs before making edits helps avoid unnecessary complexity and maintains codebase integrity.

Can I use these coding guidelines for both writing and reviewing workflows?

Yes, you can use these coding guidelines for both writing and reviewing workflows. They enforce explicit thought processes and minimal, surgical changes to surface assumptions and verify results across all coding tasks.

When should I not use a minimal surgical changes approach for code edits?

Avoid a minimal surgical changes approach when a coding task requires large architectural overhauls rather than targeted edits. These guidelines focus on reducing complexity through plan-driven execution and verifiable success criteria for minimal modifications.