skills.rest
friends-of-ute-valley-park
friends-of-ute-valley-parkOFFICIAL
Explore all skills by this author →
GitHub

karpathy-guidelines

Guide coding tasks with assumptions, tradeoffs, and verifiable success criteria.

Updated Oct 19, 2020
Search Tags:#code-review#coding-best-practices#karpathy-guidelines#LLM-coding#surgical-edits#think-before-code#verifiable-criteria
One-click install
npx skills add https://github.com/friends-of-ute-valley-park/friends-of-ute-valley --skill karpathy-guidelines-friends-of-ute-valley-park
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: karpathy-guidelines
Source: https://github.com/friends-of-ute-valley-park/friends-of-ute-valley/tree/main/.gemini/skills/karpathy-guidelines
Command: npx skills add https://github.com/friends-of-ute-valley-park/friends-of-ute-valley --skill karpathy-guidelines-friends-of-ute-valley-park

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides behavioral guidelines to reduce common LLM coding mistakes, drawing from Karpathy's observations to help writers, reviewers, and refactors keep code simple and verifiable.

Core Features & Use Cases

  • Guidance to surface assumptions during coding tasks.
  • Emphasis on surgical changes and minimal, well-scoped edits.
  • Use Case: Apply the guidelines during writing, reviewing, or refactoring AI-assisted code to avoid overcomplication and surface verifiable success criteria.

Quick Start

Start a coding task by explicitly listing assumptions, surfacing tradeoffs, and proposing alternatives; then state concrete success criteria. During reviews, identify simple, surgical changes and avoid unnecessary abstractions. Follow the three rules: Think Before Coding, Simplicity First, and Surgical Changes.

Frequently Asked Questions about karpathy-guidelines

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

FAQPage Schema
How do I reduce common LLM coding mistakes during code generation and refactoring?▼

Surgical code edits require thinking before coding, prioritizing simplicity, and making minimal, well-scoped changes. This approach avoids unnecessary abstractions and ensures modifications are evidence-based, directly addressing the specific goal without overcomplicating the codebase.

What is the best way to define verifiable success criteria for AI-assisted code?▼

Defining verifiable success criteria involves explicitly listing assumptions, surfacing tradeoffs, and proposing alternatives before coding begins. This structured decision-making process ensures AI-assisted code remains simple, goal-driven, and verifiable throughout the development lifecycle.

Can I apply these coding best practices to any LLM coding workflow?▼

Yes, you can apply these coding best practices to any LLM writing, reviewing, or refactoring workflow. The guidelines are designed to integrate into your existing process, helping you surface assumptions and enforce surgical changes regardless of the specific AI model used.

Why should I think before coding when using AI for software engineering tasks?▼

Thinking before coding is crucial because it enforces structured decision-making and surfaces hidden assumptions early. By prioritizing simplicity and planning explicit, verifiable criteria, you prevent overcomplication and ensure the LLM generates goal-driven, surgical code edits.

Related Skills

Explore More Software Engineering Skills→
kenlckkenlck
5

karpathy-guidelines

Apply Karpathy-style behavioral guidelines to code writing, review, and refactoring.

Community
Advanced
louderthanthunderx1louderthanthunderx1

karpathy-guidelines

Provide decision-making guidelines for LLM-assisted coding tasks.

Community
Intermediate
endorphin-aiendorphin-ai
4

four-principles

Enforces four engineering principles for coding in software development workflows.

Official
Intermediate
ZiadNagarZiadNagar

karpathy-guidelines

Apply Karpathy-style guidelines to coding tasks for explicit assumptions and verifiable success criteria.

Community
Intermediate