6 weeks of Claude Code

Really agree with the author's thoughts on maintenance here. I've run into a ton of cases where I would have written a TODO or made a ticket to capture some refactoring and instead just knocked it out right then with Claude. I've also used Claude to quickly try out a refactoring idea and then abandoned it because I didn't like how it came out. It really lowers the activation energy for these kinds of maintenance things.

Letting Claude rest was a great point in the article, too. I easily get manifold value compared to what I pay, so I haven't got it grinding on its own on a bunch of things in parallel and offline. I think it could quickly be an accelerator for burnout and cruft if you aren't careful, so I keep to a supervised-by-human mode.

Wrote up some more thoughts a few weeks ago at https://www.modulecollective.com/posts/agent-assisted-coding....

As someone who is programming computers for almost 30 years and professionally for about 20 by all means do some of it manually, but leverage LLMs in tutor/coach mode, with „explain this but don’t solve it for me” prompts when stuck. Let the tool convert the boring parts once you’re confident they’re truly boring.

Programming takes experience to acquire taste for what’s right, what’s not, and what smells bad and will bite you but you can temporarily (yeah) not care. If you let the tool do everything for you you won’t ever acquire that skill, and it’s critical to judge and review your work and work of others, including LLM slop.

I agree it’s hard and I feel lucky for never having to make the LLM vs manual labor choice. Nowadays it’s yet another step in learning the craft, but the timing is wrong for juniors - you are now expected to do senior level work (code reviews) from day 1. Tough!

I think this is a really interesting point. I have a few thoughts as a read it (as a bit of a grey-beard).

Things are moving fast at the moment, but I think it feels even faster because of how slowly things have been moving for the last decade. I was getting into web development in the mid-to-late-90s, and I think the landscape felt similar then. Plugged-in people kinda knew the web was going to be huge, but on some level we also know that things were going to change fast. Whatever we learnt would soon fall by the wayside and become compost for the next new thing we had to learn.

It certainly feels to me like things have really been much more stable for the last 10-15 years (YMMV).

So I guess what I'm saying is: yeah, this is actually kinda getting back to normal. At least that is how I see it, if I'm in an excitable optimistic mood.

I'd say pick something and do it. It may become brain-compost, but I think a good deep layer of compost is what will turn you into a senior developer. Hopefully that metaphor isn't too stretched!

I think Claude Code is great, but I really grew accustomed to the "Cursor-tab tab tab" autocomplete style. A little perplexed why the Claude Code integration into VS Code doesn't add something like this? It would make it the perfect product to me. Surprised more people do not talk about this/it isn't a more commonly requested feature.

Claude Code is ahead of anything else, in a very noticeable way. (I've been writing my own cli tooling for AI codegen from 2023 - and in that journey I've tried most of the options out there. It has been a big part of my work - so that's how I know.)

I agree with many things that the author is doing:

1. Monorepos can save time

2. Start with a good spec. Spend enough time on the spec. You can get AI to write most of the spec for you, if you provide a good outline.

3. Make sure you have tests from the beginning. This is the most important part. Tests (along with good specs) are how an AI agent can recurse into a good solution. TDD is back.

4. Types help (a lot!). Linters help as well. These are guard rails.

5. Put external documentation inside project docs, for example in docs/external-deps.

6. And finally, like every tool it takes time to figure out a technique that works best for you. It's arguably easier than it was (especially with Claude Code), but there's still stuff to learn. Everyone I know has a slightly different workflow - so it's a bit like coding.

I vibe coded quite a lot this week. Among them, Permiso [1] - a super simple GraphQL RBAC server. It's nowhere close to best tested and reviewed, but can be quite useful already if you want something simple (and can wait until it's reviewed.)

[1]: https://github.com/codespin-ai/permiso