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

Feature Request: Support AGENTS.md

github.com

Post reply on HN
Print view View original

231–236 of 236 posts

Re: Feature Request: Support AGENTS.md

#231

Post by monksy »

dom96 wrote: post #153
spaqin wrote: post #93

Earlier quoted context omitted.

Scrapers won't be stopped by needing an account; in the recent discussion [0] about this, someone mentioned that there's even a JSON API you can use - by appending .json to any post! It's to drive engagement... for the advertisers. They gotta keep the valuation up after the IPO. [0] https://news.ycombinator.com/item?id=49015533

I exposed this JSON API in the same style as their original API at api.reddiw.com and they banned my account without warning. I explicitly included my reddit username in the user-agent so that they could reach out if they had a problem with it. Nope, straight to banning my decade old account. The company has turned evil sadly.

spez likes your comment.

Re: Feature Request: Support AGENTS.md

#232

Post by weplay0628 »

superfrank wrote: post #49
threecheese wrote: post #39

Earlier quoted context omitted.

Are you saying this wouldn’t happen if the file had a different name? Or just that codex is better at instruction following.

No. I'm saying that the pattern of having one instruction file that just says "read this other file" doesn't seem to work well. Having a CLAUDE.md that just says "Read AGENTS.md" resulted in Claude randomly not following the rules. We tried the other way and it didn't seem like it was any better, but also, given that AGENTS.md is the standard everywhere except for Claude Code, I don't really want CLAUDE.md to be the…

[flagged]

Re: Feature Request: Support AGENTS.md

#233

Post by vrighter »

brookst wrote: post #209
vrighter wrote: post #152

Earlier quoted context omitted.

just the filename. The contents of the file are unstructured, besides being markdown

But the semantic meaning can be agent-specific. For instance, my Claude.md files have things like “never present multiple questions in walls of text, use `AskUserQuestion` with clear tradeoffs described”

it is still just an unstructured file

Re: Feature Request: Support AGENTS.md

#234

Post by DannyBee »

olejorgenb wrote: post #229
DannyBee wrote: post #218

Earlier quoted context omitted.

If i'm trying to steel-man why, I presume because the read/write/edit tools use more context tokens because they don't support reading part of a file/etc. So the agent is going to put less into context when it uses sed to see 15 lines of a file than using read and putting the entire file into context. That is my best charitable guess at what they are hoping to achieve. Of course, there is an obvious set of solutions…

Surely the read and update tools support ranges

Read does, but it rarely uses it.

Part of the reason i assume is that to extract 5 line ranges is a single line very small sed call that it has to output, but 5 separate full read calls it has to output correctly and quickly.

Who knows.

Post reply on HN