I've been using a project called SX https://github.com/sleuth-io/sx to manage them. It has its quirks, but overall I find it to be very useful.

One thing I've had to write as a layer on top of it is a way to assemble agent-specific CLAUDE.md / AGENTS.md from fragments.

For example, I have a little fragment that has all agents respond to me in ordered list format. (Since they often ask a bunch of question all jumbled throughout a response, the ordered list format allows me to respond to those specific questions.) And I also have a growing anti-Claudeism fragment as well.

Then I combine this with project-specific fragments and have it assembled into into a single CLAUDE.md / AGENTS.md. The layer also does a little reporting on the length of the resulting files and notifies me if it ever grows beyond a certain size.

"If you have a problem, that means it can be solved. If it cannot be solved, then it is not a problem. It is a reality to be accepted.” — Gus Fring, Breaking Bad

I don’t mean this callously. People confuse not having a choice with not liking the consequences of their choices.

You always have a choice. At minimum you can always choose to accept your circumstances. If you choose to not accept them, then your options are to choose to change them and accept the consequences of that choice, or to not change them and in the process choose to be a victim of circumstance and live in a state of complaint or dis-ease or anger or whatever other ill you create for yourself.

Those are literally your only options. The circumstances are seemingly not going to change. That means you have to. Whatever choice you are making now is seemingly not serving you, ergo a post on HN. What new choices can you make that will lead to a different result?

Agreed, and then the actual text of the post gets worse. It’s actually unreadable; I literally could not follow what the article was about (without putting in the effort to unravel the Claudisms, a little like unraveling Shakespeare or Finnegan’s Wake) and gave up.

It seems we are all (on HN) thoroughly convinced that the models have all gotten worse at writing as the version numbers have gone up and they are supposedly more powerful models. Sol 5.6 is nowhere near as bad as Opus 5 or even the 4 series but not great.

But looking back there was a time we were enamored with how they wrote. Perhaps it’s related to the newfound prohibitions on “write in this writer’s style”? I wonder if training the models to not be able to do that anymore has steered them towards not being able to write, at all, in any existing human good writing style and they are navigating the small pathway through all existing good human writing to whatever you want to call what they are doing now?

I'm willing to bet it's because model providers have gone all-in on programming and reasoning ability when training new models. It's where the money is, if nothing else, and the writing quality will have suffered as a knock on effect.

Someone should build a capable technical writing model, would make a TON of money

It was the newness and surprise that the quality was as good as it was, just like images from Stable Diffusion, etc. Now that we are used to it, we notice the uncanny valley speak.

> newfound prohibitions on “write in this writer’s style”

that explains a lot. but do you have a decent citation?

“Every tool is scar tissue over a real wound. Follow the wounds, and the map draws itself. THE ONE IDEA THAT UNLOCKS EVERYTHING” uh-huh

I have noticed myself having a visceral reaction to the AI tells now, somewhere between disgust and anger. And I am not an angry man and I like AI! There is just something about the voice that is incredibly off-putting, like a know-it-all friend who just ripped a giant fart. It is annoyingly correct and also smells really bad.

I don't know how else to put it, but it has this built in condescension that rides along when someone is hitting you with too many clever metaphors. Like it's TED talking at you or something, it just frames everything in a tone of authoritative/lecturing style. Very off putting just as if an actual person did it you would feel irritated by them

It sounds like an overly earnest TEDx talker leaning over you talking about something inconsequential as if it’s the most mic-droppingly insightful thing in the world.

“the matchup feels earned” is a current AI-written tell. To whom does it feel earned? To the AI that wrote this article?

I don’t know what it is specifically, but my weak human pattern-matching skills find this kind of language increasingly revolting. I don’t know why it is revolting, per se. It’s just the feeling I get.

Of course, me saying this on HN will get incorporated into GPT-5.6.175 or Claude 4.93 and it will make some version that just moves the revolting frontier elsewhere…

I think it's because it's using storytelling-like language to describe reality.

"Harry finally had control of the broom. Draco was dead in his sights. The matchup feels earned."

It's because they assume you know what precision is in regards to this comparison. Normal people don't use such words.

The same is true of humans, not surprisingly.

If you ask the average human "Why?", they will generally get defensive, especially if you are asking them to justify their own motivation.

However, if you ask them to describe the thinking and actions that led to their result, they often respond very differently.

Gender is mutable. Sex is determined by genetics, though in a handful of creatures, sex could be considered mutable.

You're right, but there was an article on here in the last week about genetic therapy correcting Down Syndrome.

I wonder how long before sex does actually become mutable.

You can change sex

And people can be born intersex

Your data is wrong

The fact that some people can be born with 6 fingers doesn't mean that saying "a human hand has 5 fingers" is generally false.

I the sense of what OP did say, it would be more "grafting a sixth finger doesn't mean that you are not human anymore".

Just like taking meat from your arm to create new genitals doesn't make you a man, biologically (psychologically it may be different).

Yeah this is a good one. I’ll add it. It really is so bizarre, it takes like three unlocks.

As the list of bugs grows, it would be helpful to have some sort of floating table of contents / quick navigation component to quickly navigate between bugs - just an idea :)

I thought about adding that bug but it seemed like “search doesn’t work in this other thing” was just going to get repetitive (despite being true and annoying AF) and I had already spent way more time on it than I should have. Put in a pull request, tho. Happy to add it!

Huh. How else would you get a massive company that has let certain bugs sit for decades to address them? This method is probably just as ineffective as reporting another bug into the Black Hole of Bug Reports that is Apple, but this one is at least mildly funnier? YMMV of course.

The search bugs in Mail are absolutely real. They might not affect every user but I've had major issues on both macOS and iOS mail, and eventually gave up and switched to Gmail.

Have you checked if it’s the fault of your mail providers? The search function queries the remote mail servers in addition to the local cache. If the mail servers return garbage you get garbage results.

This was with an iCloud email address.

Maybe iCloud search is rubbish.

Mail search works flawlessly for me with Gmail and Outlook. It takes a sec but the results are always relevant.

As I used to ask field teams and PMs back when I was a SWE - provide the steps to reproduce the bug as I can't seem to recreate it.

I experience at least five of those bugs daily and it makes me so frustrated: Hotspot, airdrop, mail search, corner drag, text selection.