Karpathy’s Claude.md turns into a practical playbook for coding with LLMs
Roger_M_Taylor · x · 2026-07-24
Andrej Karpathy’s Claude.md note is being shared as a practical coding workflow for getting language models to write code you won’t need to rewrite.
- The note collects field-tested rules for LLM-assisted programming: read before writing, think before coding, simplify aggressively, make surgical changes, verify with tests, and debug from full traces rather than guessing.
- The shared image frames the document as a compact playbook for reducing the gap between code that merely runs and code that is actually maintainable.
- Because it is tied to Anthropic’s Claude workflow and framed as a reusable engineering note, the post is more than a personality mention: it is a concrete agent/coding-method artifact.
Related event: Karpathy's Coding Insights Turned into Open-Source Claude Rules(4 posts)→
More from coding & agent
- Dev Uses Claude to Build 3D City Generator and Weather Engine — repligate · 2026-07-24
- Mnemos shifts away from full agent orchestration and doubles down on identity — RileyRalmuto · 2026-07-24
- Codex helps build a route-planning web app for 100+ bar shifts — gabrielchua · 2026-07-24
- Long-Horizon Terminal-Bench leaderboard adds a new agent eval for terminal code tasks — Muennighoff · 2026-07-24
- GPT-5.6 Sol is shown unfollowing non-human accounts directly in Chrome — cocktailpeanut · 2026-07-24
- Open-source tax engine hits 96% on TaxCalcBench, topping Fable 5 and SOL — Intelligent_Prompt18 · 2026-07-24