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

The creator of Claude Code's Claude setup

twitter.com

Post reply on HN
Print view View original

171–180 of 423 posts

Re: The creator of Claude Code's Claude setup

#171

Post by llmslave2 »

roncesvalles wrote: post #152
llmslave2 wrote: post #28

Couple things stand out to me: 1) everyone on the team uses Claude code differently. 2) Claude Code has been around for almost a year and is being built by an entire team, yet doesn't seem to have benefited from this approach. The program is becoming buggier and less reliable over time, and development speed seems indistinguishable from anything else. 3) Everything this person says should be taken with a massive grai…

>Everything this person says should be taken with a massive grain of salt considering their various conflicts of interest. Exactly, he has to dogfood it. He can't just say "actually this is a massively annoying way of developing software and probably slows me down".

He also has very clear incentives to not only promote it, but believe in it.

Re: The creator of Claude Code's Claude setup

#172

Post by vidarh »

eru wrote: post #133
AtlasBarfed wrote: post #132

Earlier quoted context omitted.

Claude is absolutely plastering Facebook with this bullshit. Every PR Claude makes needs to be reviewed. Every single one. So great! You have 10 instances of Claude doing things. Great! You're still going to need to do 10 reviews.

That's why you have Codex review the code. (I'm only half joking. Having one LLM review the PRs of another is actually useful as a first line filter.)

Even having Opus review code written by Opus works very well as a first pass. I typically have it run a sub-agent to review its own code using a separate prompt. The sub-agents gets fresh context, so it won't get "poisoned" by the top level contexts justifications for the questionable choices it might have made. The prompts then direct the top level instance to repeat the verification step until the sub-agent gives the code a "pass", and fix any issues flagged.

The result is change sets that still need review - and fixes - but are vastly cleaner than if you review the first output.

Doing runs with other models entirely is also good - they will often identify different issues - but you can get far with sub-agents and different persona (and you can, if you like, have Claude Code use a sub agent to run codex to prompt it for a review, or vice versa - a number of the CLI tools seems to have "standardized" on "-p " to ask a question on the command line)

Basically, reviewing output from Claude (or Codex, or any model) that hasn't been through multiple automated review passes by a model first is a waste of time - it's like reviewing the first draft from a slightly sloppy and overly self-confident developer who hasn't bothered checking if their own work even compiles first.

Re: The creator of Claude Code's Claude setup

#173

Post by allan_s »

yoz-y wrote: post #156
exitb wrote: post #107

Earlier quoted context omitted.

Multiple instances of agents are an equivalent to tabs in other applications - primarily holders of state, rather than means for extreme parallelism.

I have not used Claude. But my experience with Gemini and aider is that multiple instances of agents will absolutely stomp over each other. Even in a single sessions overwriting my changes after telling the agent that I did modifications will often result in clobbering.

See the agent as a coworker ssh-ing on your machine, how would you work efficiently ? By working on the same directory ? No

You give each agent a git worktree and if you want to check, you checkout their branch.

Re: The creator of Claude Code's Claude setup

#174

Post by djmips »

erichocean wrote: post #165
stingraycharles wrote: post #128

Earlier quoted context omitted.

I hope self-promotion isn't frowned upon, but I've been spending the past months figuring out a workflow [1] that helps tackle the "more complicated problems" and ensure long-term maintainability of projects when done purely through Claude Code. Effectively, I try to: - Do not allow the LLM to make any implicit decisions, but instead confirm with the user. - Ensure code is written in such a way that it's easy to unde…

> Ensure code is written in such a way that it's easy to understand for LLMs Over the summer last year, I had the AI (Gemini Pro 2.5) write base libraries from scratch that area easy for itself to write code against. Now GPro3 can one-shot (with, at most, a single debug loop at the REPL) 100% of the normal code I need developed (back office/business-type code). Huge productivity booster, there are a few things that a…

How would you characterize code is easy for AI to write code against. - and wouldn't that also be true for humans?

Re: The creator of Claude Code's Claude setup

#175

Post by djmips »

jedberg wrote: post #30

The funniest part of that whole thing was when someone said "I trusted you, but you use light mode on your terminal" and then he replied that people stop by his desk daily just to make fun of him for it.

I started with dark mode because that's what amber text on a terminal was... but then the big thing was UI's simulating paper and then we had a turn to dark mode but recently I've gone back to the light side.

Re: The creator of Claude Code's Claude setup

#176

Post by ChadNauseam »

crystal_revenge wrote: post #169
tmerr wrote: post #65

This is interesting to hear, but I don't understand how this workflow actually works. I don't need 10 parallel agents making 50-100 PRs a week, I need 1 agent that successfully solves the most important problem. I don't understand how you can generate requirements quicky enough to have 10 parallel agents chewing away at meaningful work. I don't understand how you can have any meaningful supervising role over 10 thing…

My initial response to reading this post was "wow, I think I'd rather just write the code". I also remain a bit skeptical because, if all of this really worked (and I mean over a long time and scaling to meet a range of business requirements), even if it's not how I personally want to write code, shouldn't we be seeing a ton of 1 person startups? I see Bay area startups pushing 996 and requiring living in the Bay are…

I do stuff in my free time now that would have been a full time job a year ago. Accomplishing in months what would have taken years. (And doing in days what would have taken weeks.) I'm talking about actually built-out products with a decent amount of code and features, not basic prototypes. I feel like the vibe is "put up or shut up", so check out my bio for one example.

I think your logic goes wrong because you assume that more productivity implies less desire for engineers. But now engineers are maybe 2x or 5x more productive than before. So that makes them more attractive to hire than before. It's not like there was some fixed pool of work to be done and you just had to hire enough to exhaust the pool. It's like if new pickaxes were invented that let your gold miners dig 5x more gold. You'd see an explosion in gold miners, not a reduction. For another example, I spend all my free time coding now because I can do so much now. I get so much more result for the same effort, that it makes sense to put more effort in.

Re: The creator of Claude Code's Claude setup

#177

Post by webdev1234568 »

rdiddly wrote: post #157

I feel like it's time for me to hang up this career. Prompting is boring, and doing it 5 times at once is just annoying multitasking. I know I'm mostly in it for the money, but at least there used to be a feeling of accomplishment sometimes. Now it's like, whose accomplishment is it?

Don't give up to the facade just yet.

This is the creator of a product saying how good it is.

If you've worked anywhere professionally you know how every place has its problems, where people just lie constantly about things?

Yeah.

Keep at it and see where things go.

I'm also a dev a bit overwhelmed by all of this talk, at my job I've tried quite a few things and I'm still mostly just using copilot for auto complete and very small tasks that I review throughly, everything else is manually.

If this is indeed the future I also don't wanna be a part of it and will switch to another career, but all this talk seems to come only from the people who actually built these things.

Re: The creator of Claude Code's Claude setup

#178

Post by wiseowise »

llmslave2 wrote: post #28

Couple things stand out to me: 1) everyone on the team uses Claude code differently. 2) Claude Code has been around for almost a year and is being built by an entire team, yet doesn't seem to have benefited from this approach. The program is becoming buggier and less reliable over time, and development speed seems indistinguishable from anything else. 3) Everything this person says should be taken with a massive grai…

> Claude Code has been around for almost a year and is being built by an entire team, yet doesn't seem to have benefited from this approach. The program is becoming buggier and less reliable over time, and development speed seems indistinguishable from anything else.

Shhh, this is not what you’re supposed to look at.

Look! Bazillion more agents! Gorrilion more agents! Productivity! Fire those lazy code monkeys, buy our product! Make me riiiich.

Re: The creator of Claude Code's Claude setup

#179

Post by llmslave2 »

crystal_revenge wrote: post #169
tmerr wrote: post #65

This is interesting to hear, but I don't understand how this workflow actually works. I don't need 10 parallel agents making 50-100 PRs a week, I need 1 agent that successfully solves the most important problem. I don't understand how you can generate requirements quicky enough to have 10 parallel agents chewing away at meaningful work. I don't understand how you can have any meaningful supervising role over 10 thing…

My initial response to reading this post was "wow, I think I'd rather just write the code". I also remain a bit skeptical because, if all of this really worked (and I mean over a long time and scaling to meet a range of business requirements), even if it's not how I personally want to write code, shouldn't we be seeing a ton of 1 person startups? I see Bay area startups pushing 996 and requiring living in the Bay are…

And to push this example further, if you can hire 10 developers each commanding 10 reliable junior-mid developers you have a team of 100, which is probably more than enough to build basically any software project in existence. WhatsApp was built with way less than that.

Re: The creator of Claude Code's Claude setup

#180

Post by sneak »

crystal_revenge wrote: post #169
tmerr wrote: post #65

This is interesting to hear, but I don't understand how this workflow actually works. I don't need 10 parallel agents making 50-100 PRs a week, I need 1 agent that successfully solves the most important problem. I don't understand how you can generate requirements quicky enough to have 10 parallel agents chewing away at meaningful work. I don't understand how you can have any meaningful supervising role over 10 thing…

My initial response to reading this post was "wow, I think I'd rather just write the code". I also remain a bit skeptical because, if all of this really worked (and I mean over a long time and scaling to meet a range of business requirements), even if it's not how I personally want to write code, shouldn't we be seeing a ton of 1 person startups? I see Bay area startups pushing 996 and requiring living in the Bay are…

A lot of people either a) don’t know about the good tools or b) aren’t using them enough/properly.

There is a ton of anti-AI sentiment, and not all LLMs are equal. There is a lot of individual adoption that is yet to occur.

I know at least two startups that are one person or two people that are punching way above their weight due to this force multiplier. I don’t think it’s industry-wide yet, but it will be relatively soon.

Check back in on your assessment in a year.

Post reply on HN