skills.rest
Takfes
TakfesCOMMUNITY
Explore all skills by this author →
GitHub

karpathy-guidelines

Enforce explicit assumptions, simpler solutions, and surgical edits in code tasks.

Updated May 6, 2026
Search Tags:#code review#refactoring#llm#assumptions#simplicity#coding guidelines
One-click install
npx skills add https://github.com/Takfes/indie-scaffolder --skill karpathy-guidelines-takfes
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: karpathy-guidelines
Source: https://github.com/Takfes/indie-scaffolder/tree/main/components/agent-skills-commands/.agents/skills/karpathy-guidelines
Command: npx skills add https://github.com/Takfes/indie-scaffolder --skill karpathy-guidelines-takfes

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps reduce common AI coding mistakes by encouraging explicit assumptions, simpler solutions, and careful changes instead of overengineering or guesswork.

Core Features & Use Cases

  • Assumption Surfacing: Encourages the model to state uncertainty, compare interpretations, and ask clarifying questions before acting.
  • Simplicity First: Pushes for the smallest code change that solves the request without speculative abstractions or unnecessary features.
  • Surgical Editing: Keeps edits tightly scoped so existing code, style, and unrelated logic remain untouched.
  • Verifiable Progress: Promotes brief plans and success criteria so implementation can be checked step by step.
  • Use Case: Ideal for code generation, debugging, refactoring, and review when you want the assistant to stay precise, conservative, and test-driven.

Quick Start

Ask the assistant to apply the Karpathy Guidelines while writing, reviewing, or refactoring your code changes.

Frequently Asked Questions about karpathy-guidelines

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

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

To stop LLM coding mistakes, enforce explicit assumptions, simpler solutions, and surgical edits. This forces the model to surface uncertainty, ask clarifying questions, and apply minimal changes to preserve existing behavior.

What is the best way to ensure LLM code review preserves existing behavior?▼

The best way to preserve existing behavior is using surgical editing guidelines that keep changes tightly scoped. This prevents unrelated logic and code style from being touched during reviews.

Why does my LLM add unnecessary features and speculative abstractions to simple code changes?▼

LLMs add speculative abstractions because they lack enforced simplicity constraints. Applying a simplicity-first guideline pushes for the smallest code change that solves the request without overengineering.

How do I apply explicit assumptions and verifiable success criteria to LLM refactoring?▼

To apply explicit assumptions, prompt the LLM to state uncertainties and compare interpretations before acting. This creates verifiable success criteria so implementation can be checked step by step.

When should I use surgical editing guidelines for LLM code generation?▼

Use surgical editing guidelines when clarity is limited, multiple interpretations exist, or existing behavior must be preserved. This ensures minimal changes and prevents the model from overwriting unrelated logic.

Related Skills

Explore More Software Engineering Skills→
VKirillVKirill
15

karpathy-guidelines

Enforces assumption-aware, minimal, verified coding changes for LLMs.

Community
Basic
vanducngvanducng
5

simplify

Simplify nested conditionals and long functions without changing behavior.

Community
Intermediate
Nero-kkNero-kk

karpathy-guidelines

Enforce explicit assumptions and minimal implementations in LLM-generated code.

Community
Basic
NagyViktNagyVikt
2

prompt-optimizer

Compress natural language prompts into structured, token-efficient instructions.

Community
Intermediate