What problem does it solve? LLM coding assistants often overcomplicate solutions, make sweeping unrequested changes, hide assumptions, and declare success without verification. This Skill provides a set of behavioral guidelines, derived from Andrej Karpathy's observations on LLM coding pitfalls, that keep code changes minimal, explicit, and verifiable. ## Core Features & Use Cases - Think Before Coding: Forces explicit statement of assumptions, presentation of alternative interpretations, and stopping to ask when requirements are unclear. - Simplicity and Surgical Changes: Enforces minimum-code solutions with no speculative abstractions, and restricts edits to only what the request requires while matching existing style. - Goal-Driven Execution: Transforms vague tasks into verifiable success criteria (e.g., write a failing test, then make it pass) with step-by-step verification plans. - Use Case: When asking an AI to fix a bug or refactor a module, apply these guidelines so the agent writes a reproduction test first, changes only the relevant lines, and verifies the fix instead of self-attesting success. ## Quick Start Ask the agent to apply the karpathy-guidelines skill while implementing or reviewing your next code change so it states assumptions, keeps edits surgical, and verifies against defined success criteria.
