Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Writing a good Claude.md

humanlayer.dev

Post reply on HN
Print view View original

1–10 of 304 posts

Re: Writing a good Claude.md

#2

Post by eric-burel »

"You can investigate this yourself by putting a logging proxy between the claude code CLI and the Anthropic API using ANTHROPIC_BASE_URL" I'd be eager to read a tutorial about that I never know which tool to favour for doing that when you're not a system or network expert.

Re: Writing a good Claude.md

#3

Post by jasonjmcghee »

Interesting selection of models for the "instruction count vs. accuracy" plot. Curious when that was done and why they chose those models. How well does ChatGPT 5/5.1 (and codex/mini/nano variants), Gemini 3, Claude Haiku/Sonnet/Opus 4.5, recent grok models, Kimi 2 Thinking etc (this generation of models) do?

Re: Writing a good Claude.md

#5

Post by alansaber »

jasonjmcghee wrote: post #3

Interesting selection of models for the "instruction count vs. accuracy" plot. Curious when that was done and why they chose those models. How well does ChatGPT 5/5.1 (and codex/mini/nano variants), Gemini 3, Claude Haiku/Sonnet/Opus 4.5, recent grok models, Kimi 2 Thinking etc (this generation of models) do?

Guessing they included some smaller models just to show how they dump accuracy at smaller context sizes

Re: Writing a good Claude.md

#7

Post by fishmicrowaver »

eric-burel wrote: post #2

"You can investigate this yourself by putting a logging proxy between the claude code CLI and the Anthropic API using ANTHROPIC_BASE_URL" I'd be eager to read a tutorial about that I never know which tool to favour for doing that when you're not a system or network expert.

Have you considered just asking claude? I'd wager you'd get up and running in <10 minutes.

Re: Writing a good Claude.md

#8

Post by andersco »

I have found enabling the codebase itself to be the “Claude.md” to be most effective. In other words, set up effective automated checks for linting, type checking, unit tests etc and tell Claude to always run these before completing a task. If the agent keeps doing something you don’t like, then a linting update or an additional test often is more effective than trying to tinker with the Claude.md file. Also, ensure docs on the codebase are up to date and tell Claude to read relevant parts when working on a task and of course update the docs for each new task. YMMV but this has worked for me.

Re: Writing a good Claude.md

#9

Post by jasonjmcghee »

alansaber wrote: post #5
jasonjmcghee wrote: post #3

Interesting selection of models for the "instruction count vs. accuracy" plot. Curious when that was done and why they chose those models. How well does ChatGPT 5/5.1 (and codex/mini/nano variants), Gemini 3, Claude Haiku/Sonnet/Opus 4.5, recent grok models, Kimi 2 Thinking etc (this generation of models) do?

Guessing they included some smaller models just to show how they dump accuracy at smaller context sizes

Sure - I was more commenting that they are all > 6 months old, which sounds silly, but things have been changing fast, and instruction following is definitely an area that has been developing a lot recently. I would be surprised if accuracy drops off that hard still.

Re: Writing a good Claude.md

#10

Post by szundi »

andersco wrote: post #8

I have found enabling the codebase itself to be the “Claude.md” to be most effective. In other words, set up effective automated checks for linting, type checking, unit tests etc and tell Claude to always run these before completing a task. If the agent keeps doing something you don’t like, then a linting update or an additional test often is more effective than trying to tinker with the Claude.md file. Also, ensure…

[dead]
Post reply on HN