# Kyle Mistele - Writing a good CLAUDE.md (Highlights) ![rw-book-cover|256](https://humanlayer.dev/api/writing-a-good-claude-md/og) ## Highlights - Frontier thinking LLMs can follow ~ 150-200 instructions with reasonable consistency. ([View Highlight](https://read.readwise.io/read/01kc2t5c3jcet8dk89hnchrmet)) ^965289583 - Since `CLAUDE.md` goes into *every single session*, you should ensure that its contents are as universally applicable as possible. ([View Highlight](https://read.readwise.io/read/01kc2t7p279cc0x6mr4e5vm0s8)) ^965289665 - While Anthropic does not have an official recommendation on how long your `CLAUDE.md` file should be, general consensus is that < 300 lines is best, and shorter is even better. ([View Highlight](https://read.readwise.io/read/01kc2t83a5z64ch50796vscksc)) ^965289689 - At HumanLayer, our root `CLAUDE.md` file is *less than sixty lines*. ([View Highlight](https://read.readwise.io/read/01kc2t8bp5zcbh2k9av8kbp8zc)) ^965289694 - Instead of including all your different instructions about building your project, running tests, code conventions, or other important context in your `CLAUDE.md` file, we recommend keeping task-specific instructions in *separate markdown files* with self-descriptive names somewhere in your project. ([View Highlight](https://read.readwise.io/read/01kc2t94e58cbpdz48w0nz99p5)) ^965289715 - **Never send an LLM to do a linter's job**. LLMs are comparably expensive and *incredibly* slow compared to traditional linters and formatters. We think you should *always use deterministic tools whenever you can*. ([View Highlight](https://read.readwise.io/read/01kc2tarpw531sg1mjjqzf2we5)) ^965289928 - Don't use `/init` or auto-generate your `CLAUDE.md` ([View Highlight](https://read.readwise.io/read/01kc2tejt0npk352nc2hwaa9dr)) ^965292835 #opinion