Hacker News Reader

Comment by bloppe | Hacker News Reader

An I the only one who has to constantly tell Claude and Gemini to stop making edits to my codebase because they keep messing things up and breaking the build like ten times in a row, duplicating logic everywhere, etc? I keep hearing about how impressive agents are. I wish they could automate me out of my job faster

View on HN

Are you paying for the higher end models? Do you have proper system prompts and guidance in place for proper prompt engineering? Have you started to practice any auxiliary forms of context engineering?

This isn't a magic code genie, it's a very complicated and very powerful new tool that you need to practice using over time in order to get good results from.

It ain’t a magic code genie. And developers don’t spend most of their day typing lines of code. Lots of it is designing, figuring out what to build, understanding the code, maintenance considerations, and adhering to the style of whatever file you’re in. All these agents needing local context and still spit junk.

guy 1: I put money in the slot machine everyone says wins all the time and I lose

you: HAVE YOU PUT MORE TOKENS IN???? ARE YOU PUTTING THEM IN THE EXPENSIVE MACHINES???

super compelling argument /s

if you want to provide working examples of "prompt engineering" or "context engineering" please do but "just keep paying until the behavior is impressive" isn't winning me as a customer

it's like putting out a demo program that absolutely sucks and promising that if I pay, it'll get good. why put out the shit demo and give me this impression, then, if it sucks?

The way I ended up paying for Claude max was that I started on the cheap plan, it went well, then it wanted more money, and I paid because things were going well.

Then it ran out of money again, and I gave it even more money.

I'm in the low 4 figures a year now, and it's worth it. For a day's pay each year, I've got a junior dev who is super fast, makes good suggestions, and makes working code.

> For a day's pay each year

For anyone trying to back of the napkin at $1000 as 4-figures per year, averaged as a day salary, the baseline salary where this makes sense is about ~$260,000/yr? Is that about right lordnacho?

Yeah I thought that was a reasonable number in the ballpark. I mean, it probably makes sense to pay a lot more for it. A grand is probably well within the range where you shouldn't care about it, even if you only get a basic salary and it's a terrible year with no bonus.

And that's not saying AI tools are the real deal, either. It can be a lot less than a fully self driving dev and still be worth a significant fraction of an entry level dev.

I assume it's after tax too..

That's the beauty of the hype: Anyone who cannot replicate it, is “holding it wrong”.

Or maybe it works well in some cases and not others?

   > it's a very complicated and very powerful new tool that you need to practice using over time in order to get good results from.

Of course this is and would be expected to be true. Yet adoption of this mindset has been orders of magnitude slower than the increase in AI features and capabilities.

Create and point them to an agent.md file http://agent.md/

I'm really baffled why the coding interfaces have not implemented a locking feature for some code. It seems like an obvious feature to be able to select a section of your code and tell the agent not to modify it. This could remove a whole class of problems where the agent tries to change tests to match the code or removes key functionality.

One could even imagine going a step further and having a confidence level associated with different parts of the code, that would help the LLM concentrate changes on the areas that you're less sure about.

Why are engineers so obstinate about this stuff? You really need a GUI built for you in order to do this? You can't take the time to just type up this instruction to the LLM? Do you realize that's possible? You can just write instructions "Don't modify XYZ.ts file under any circumstances". Not to mention all the tools have simple hotkeys to dismiss changes for an entire file with the press of a button if you really want to ignore changes to a file or whatever. In Cursor you can literally select a block of text and press a hotkey to "highlight" that code to the LLM in the chat, and you could absolutely tell it "READ BUT DON'T TOUCH THIS CODE" or something, directly tied to specific lines of code, literally the feature you are describing. BUT, you have to work with the LLM and tooling, it's not just going to be a button for you or something.

You can also literally do exactly what you said with "going a step further".

Open Claude Code, run `/init`. Download Superwhisper, open a new file at project root called BRAIN_DUMP.md, put your cursor in the file, activate Superwhisper, talk in stream of consciousness-style about all the parts of the code and your own confidence level, with any details you want to include. Go to your LLM chat, tell it to "Read file @BRAIN_DUMP.md" and organize all the contents into your own new file CODE_CONFIDENCE.md. Tell it to list the parts of the code base and give it's best assessment of the developer's confidence in that part of the code, given the details and tone in the brain dump for each part. Delete the brain dump file if you want. Now you literally have what you asked for, an "index" of sorts for your LLM that tells it the parts of the codebase and developer confidence/stability/etc. Now you can just refer to that file in your project prompting.

Please, everyone, for the love of god, just start prompting. Instead of posting on hacker news or reddit about your skepticism, literally talk to the LLM about it and ask it questions, it can help you work through almost any of this stuff people rant about.

_all_ models I’ve tried continuously, and still, have problems ignoring rules. I’m actually quite shocked someone would write this if you have experience in the area, as it so clearly contrasts with my own experience.

Despite explicit instructions in all sorts of rules and .md’s, the models still make changes where they should not. When caught they innocently say ”you’re right I shouldn’t have done that as it directly goes against your rule of <x>”.

Just to be clear, are you suggesting that currently, with your existing setup, the AI’s always follow your instructions in your rules and prompts? If so, I want your rules please. If not, I don’t understand why you would diss a solution which aims to hardcode away some of the llm prompt interpretation problems that exist

I am by no means an AI skeptic. It is possible to encode all sorts of things into instructions, but I don’t think the future of programming is every individual constructing and managing artisan prompts. There are surely some new paradigms to be discovered here. A code locking interface seems like an interesting one to explore. I’m sure there are others.

Or, you know, chmod -w XYZ.ts

They need "context engineering" which what I'll describe best as "railing" them in. If you give them a bit of a loose space, they'll massacre your code base. You can use their freedom for exploration but not for implementation.

In essence, you have to do the "engineering" part of the app and they can write the code pretty fast for you. They can help you in the engineering part, but you still need to be able to weigh in whatever crap they recommend and adjust accordingly.

This is way even as a paid user I stick to the browser tab llms. I am a context control freak and constantly just grabbing a new sessions and starting over. I don't try and fix a session and the incentives of a subscription vs api token payment model has inverse incentives.

It happens to me, yes. Sometimes they get stuck in the process. I learned how to go around certain issues but it's very annoying.

Duplicate logic is definitely a thing. That and littering comments all over the place.

Worth it to me as I can fix all the above after the fact.

Just annoying haha

Ask for no comments. Or extremely infrequent ones for complex sections only. Agreeing with https://news.ycombinator.com/item?id=44619462 here.

Did you tell them to not duplicate code?

Reply on news.ycombinator.com