Live data from Hacker News

Feature Request: Support AGENTS.md

github.com

191–200 of 236 posts

Re: Feature Request: Support AGENTS.md

#191
post #184

Earlier quoted context omitted.

Making your tool less compatible with the rest of the ecosystem at large not only makes it harder to move to your tool, but also harder to move away from your tool. Given these AI labs basically have no moat, seems they're trying to hold on to every little piece they can do make it harder to move away, even if it's a really tiny and marginal feature, like specifically titled file on disk... Kind of embarrassing overa…

> Making your tool less compatible with the rest of the ecosystem at large not only makes it harder to move to your tool, but also harder to move away from your tool. That's true but this is really not something that's a barrier to switching, just an annoyance for users. All it takes is a rename/symlink/reference whatever from the user's perspective. If it required a lot of rearchitecting on Anthropic's side or it wa…

Depends on how you use these harnesses, I don't interact with any of them directly for example, but have my own harness around them, mostly so I can collect all the different ones under one "roof". With that, comes certain expectations of where things live. Anything that has "CC uses file path X" for example would require refactoring/changes, compared to other agents. Contrived example, but sometimes it's more than just "mv AGENTS.md CLAUDE.md".

Re: Feature Request: Support AGENTS.md

#192

I put a reference to AGENTS.md in the CLAUDE.md file, and then put Claude specific stuff (mainly info about hooks and scripts) after that. https://github.com/kstenerud/yoloai/blob/main/CLAUDE.md

I just have one as a symlink to the other. Seems to work, and doesn't need an extra tool call.

Re: Feature Request: Support AGENTS.md

#193

So many developer-hostile things lately. Just a few days ago they turned on an experiment the forces claude code to use bash over standard tools in auto mode. From the system prompt directly, new as of August 18th: Do your work through the Bash tool wherever it can accomplish the job: read files with cat, head, or sed -n, search with grep and find, and make file changes with sed, heredocs, or short scripts, rather th…

> So many developer-hostile things lately.

It is extraordinary how they managed to fuck up all of their goodwill with all these unnecessary stuff. They truly are a hostile company if I have ever seen one. And they had the entire developer community cheering for them a couple months ago.

I hope they fail in their mission, whatever that is. Because I'm sure it's no benefit to anyone ever.

Re: Feature Request: Support AGENTS.md

#194

Earlier quoted context omitted.

OK, this argument is getting semantic and, therefore, uninteresting. I think you know what I meant, even if I used some words in a different way than you would have liked.

> this argument is getting semantic Well, it was semantic from the beginning, the whole discussion I raised is because you seem unable to understand what "structured data" actually means in reality. I know perfectly fine what you meant, and you were misinformed about the meaning, it's OK to just say "Yeah, you're right" or just don't reply at all instead of double-down on not understand the meaning of words. Anyways,…

This used to be a nice place to have sensible discussions. Sadly, it seems, some people want to turn it into Reddit. Please enjoy your day, too; it seems like you could do with your mood improving.

Re: Feature Request: Support AGENTS.md

#195
Really strange; anyway it is closed source and best not to have too much reliance and attachment to any one tool. I have switched to GPT 5.5 and 5.6 Sol and Codex ove Claude already. Fable is very expensive/token limiting. I am an experienced programmer heavily using AI for many POCs now. At one time Anthropic and its models Opus,Fable looked leading; now OpenAI has caught up. I find 5.5 models more than enough for most tasks and have shifted back to codex

Re: Feature Request: Support AGENTS.md

#196

Earlier quoted context omitted.

> this argument is getting semantic Well, it was semantic from the beginning, the whole discussion I raised is because you seem unable to understand what "structured data" actually means in reality. I know perfectly fine what you meant, and you were misinformed about the meaning, it's OK to just say "Yeah, you're right" or just don't reply at all instead of double-down on not understand the meaning of words. Anyways,…

This used to be a nice place to have sensible discussions. Sadly, it seems, some people want to turn it into Reddit. Please enjoy your day, too; it seems like you could do with your mood improving.

It still is a very nice place for sensible discussions, given the right seed for discussions. People used to be thankful to learn they were wrong too, although I won't claim HN is turning into reddit, lest we break the very guidelines we apparently are upset about being broken.

Oh no, my mood is great, life couldn't be better. Don't mistake me wanting people to use words they understand for me being upset, it's just words and text, ultimately it all has little meaning compared to the grandness of life.

Re: Feature Request: Support AGENTS.md

#197

So many developer-hostile things lately. Just a few days ago they turned on an experiment the forces claude code to use bash over standard tools in auto mode. From the system prompt directly, new as of August 18th: Do your work through the Bash tool wherever it can accomplish the job: read files with cat, head, or sed -n, search with grep and find, and make file changes with sed, heredocs, or short scripts, rather th…

[flagged]

Re: Feature Request: Support AGENTS.md

#200

Earlier quoted context omitted.

Claude Code is quickly becoming an anchor. Every week I marvel at how much worse it gets, how much more essential information is hidden and replaced with bloated useless TUI and rambling tangential responses that hide the useful bits of info behind jargon invented by the agent without ever explaining it to the user. It's like the PMs for Claude Code are reward hacking their own reinforcement learning.

What tools do you find are better? I have been happy with Claude Code lately, but I haven't explored other options much in over a year. Curious to try something else out if it's less rambling and tangential.

I resist the Node.js rage by using the Rust agent https://maki.sh/ . It is very capable and has a sane plugin system via Lua.
Post reply on HN