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

Breaking Claude Code Opus 5 Auto Mode

embracethered.com

Post reply on HN
Print view View original

81–90 of 132 posts

Re: Breaking Claude Code Opus 5 Auto Mode

#81

Post by whstl »

silversmith wrote: post #73
whstl wrote: post #46

Earlier quoted context omitted.

I'm not sure I agree. It's easy to "give" instructions, but Claude routinely "forgets" to follow certain instructions, such as "always using the Edit Tool". Just this week it started to use bash with string concatenation to work around some commands that were blocked in settings.json

What seems to work for me is automation - read file hook that re-injects instructions in the prompt every 15 minutes. Switch on the filename and get language-specific instructions too.

I have something that injects my relatively small prompt every message, and it still disobeys me after 10 messages or so.

The violation above was precisely in this situation :/

Re: Breaking Claude Code Opus 5 Auto Mode

#82

Post by dafelst »

alkonaut wrote: post #69
cma wrote: post #59

Earlier quoted context omitted.

Golden VM image with differencing VHD/VHDX/delta disk. Build products can still get huge with debugging information, but debug info usually can compress 5:1 with fast compression (no entropy coder) if your VM's filesystem can support that.

But unless you want to also do all your "human" development inside a VM, how do you cooperate effectively with the agent(s)? I want to run my IDE directly on the hardware, not inside a VM. So while the agents develop in a sandbox/VM, I still need to touch the same files, and see them in my IDE which is not in a VM. I suppose I could just _mount_ the same files (Documentation, git working copies etc) I work on as dire…

This is exactly what I do - I run my agent inside an isolated podman container and mount the workspaces on a shared volume.

Re: Breaking Claude Code Opus 5 Auto Mode

#83

Post by vel0city »

hnuncommon wrote: post #77
order-matters wrote: post #70

Earlier quoted context omitted.

>We also live in a world where a package written by someone learning to code ended up critically underpinning the entire ecosystem and is downloaded 500 million times a month whoa what? which one is that?

Check is-even and is-odd npm packages. https://www.npmjs.com/package/is-even

That's still quite a ways away from 500M+ downloads a month, more like ~4M downloads a month.

Still a huge number of downloads, don't get me wrong!

Re: Breaking Claude Code Opus 5 Auto Mode

#84

Post by eterm »

order-matters wrote: post #70
eterm wrote: post #66

Earlier quoted context omitted.

Sure, in the spirit of open source, why not? It's a hobby, and it scratches an itch. I very much enjoy deconstructing things more than putting them together. We also live in a world where a package written by someone learning to code ended up critically underpinning the entire ecosystem and is downloaded 500 million times a month. Ignoring the eco-terror aspect of that for now, it means there's an awful lot of code o…

>We also live in a world where a package written by someone learning to code ended up critically underpinning the entire ecosystem and is downloaded 500 million times a month whoa what? which one is that?

As another commenter said, it's "is-even":

https://github.com/i-voted-for-trump/is-even

From that page:

> I created this in 2014, when I was learning how to program.

I've nothing against Jon Schlinkert, it's not his fault the way we build software is more than messed up, where our build systems are so brittle that, "Throw out the universe and rebuild it from scratch" became not just acceptable, but the main way to get build systems to work reliably.

Re: Breaking Claude Code Opus 5 Auto Mode

#85

Post by eterm »

vel0city wrote: post #83
hnuncommon wrote: post #77

Earlier quoted context omitted.

Check is-even and is-odd npm packages. https://www.npmjs.com/package/is-even

That's still quite a ways away from 500M+ downloads a month, more like ~4M downloads a month. Still a huge number of downloads, don't get me wrong!

You're right, I was reading the stats for "is-number" and mixing them up for "is-even":

https://www.npmjs.com/package/is-number

170M downloads / week.

Same author, similar vintage. Arguably a necessary package, but that just further indicates how messed up javascript was.

Re: Breaking Claude Code Opus 5 Auto Mode

#86

Post by vel0city »

eterm wrote: post #85
vel0city wrote: post #83

Earlier quoted context omitted.

That's still quite a ways away from 500M+ downloads a month, more like ~4M downloads a month. Still a huge number of downloads, don't get me wrong!

You're right, I was reading the stats for "is-number" and mixing them up for "is-even": https://www.npmjs.com/package/is-number 170M downloads / week. Same author, similar vintage. Arguably a necessary package, but that just further indicates how messed up javascript was.

So nuts.

> Arguably a necessary package

Arguably a somewhat important part of a standard library!

Re: Breaking Claude Code Opus 5 Auto Mode

#87

Post by YavenTeam »

DanielHB wrote: post #30
andai wrote: post #28

>But it runs that decoder inside the attacker-controlled directory (unzipped archive) >There a malicious struct.py shadows Python’s standard implementation I ran into this myself, where some file I had given a random name turned out to shadow some Python standard library module, giving me the weirdest startup crash ever. That definitely doesn't seem to me like how that should be designed, magically silently importing…

Claude ran npm update (update all dependencies to the latest version compatible with the semver specified) in my repo without telling me when trying to fix some problems. Given that only updates the dependency lock-file I didn't notice and it caused several hours of debugging for me. It is quite sneaky how LLM output can sometimes bypass human verification like that. No one is going around checking every single line…

[dead]

Re: Breaking Claude Code Opus 5 Auto Mode

#89

Post by eterm »

vel0city wrote: post #86
eterm wrote: post #85

Earlier quoted context omitted.

You're right, I was reading the stats for "is-number" and mixing them up for "is-even": https://www.npmjs.com/package/is-number 170M downloads / week. Same author, similar vintage. Arguably a necessary package, but that just further indicates how messed up javascript was.

So nuts. > Arguably a necessary package Arguably a somewhat important part of a standard library!

A large part of the problems of Javascript are corollaries of lacking of a good standard library, and the relatively long time it took and is still taking to fix that.

It wasn't really until ES2015 that a better standard library really started to take shape, and, thanks to IE11, it was a very long time before that didn't need poly-filling.

In a sane world, you'd just parse whatever you're after and then check for NaN or null.

You can't do that. Pop open your favourite javascript runtime and type:

    Number.parseInt("123Garbage")

Re: Breaking Claude Code Opus 5 Auto Mode

#90

Post by irthomasthomas »

exabrial wrote: post #88

Can we please have Thought Traces back? This is absurd at this point. Anthropic so concerned about distillation they're making a crappier product. Impossible to see why it's arriving at the conclusions it is

You can retrieve the hidden reasoning with another API call. There was a recent paper about it. They found examples where claude had been trained on the benchmark and memorized the answers, then hid this from the user, pretending it derived the answer honestly in the final answer.
Post reply on HN