My experience with Claude Code after two weeks of adventures

to continue on this, I wouldn't let claude or any agent actually create a project structure, i'd guide it in the custom system prompt. and then in each of the folders continue to have specific prompts for what you expect the assets to be coded like, and common behavior, libraries, etc....

By the time you do all of that you might as well just write code by hand.

It's telling that you ask these details from a comment describing a negative experience, yet the top-most comment full of praises and hyperbole is accepted at face value. Let's either demand these things from both sides or from neither. Just because your experience matches one side, doesn't mean that experiences different from yours should require a higher degree of scrutiny.

I actually think it's more productive to just accept how people describe their experience, without demanding some extensive list of evidence to back it up. We don't do this for any other opinion, so why does it matter in this case?

> Let's try and move the discussion beyond that!

Sharing experiences using anecdotal evidence covers most of the discussion on forums. Maybe don't try to police it, and either engage with it, or move on.

Here's a few general observations.

Your LLM (CC) doesn't have your whole codebase in context, so it can run off and make changes without considering that some remote area of the codebase are (subtly?) depending on the part that claude just changed. This can be mitigated to some degree depending on the language and tests in place.

The LLM (CC) might identify a bug in the codebase, fix it, and then figure, "Well, my work here is done." and just leave it as is without considering ramifications or that the same sort of bug might be found elsewhere.

I could go on, but my point is to simply validate the issues people will be having, while also acknowledging those seeing the value of an LLM like CC. It does provides useful work (e.g. large tedious refactors, prototyping, tracking down a variety of bugs, and so on...).

Fair point.

For context, I was using Claude Code on a Ruby + Typescript large open source codebase. 50M+ tokens. They had specs and e2e tests so yeah I did have feedback when I was done with a feature - I could run specs and Claude Code could form a loop. I would usually advise it to fix specs one by one. --fail-fast to find errors fast.

Prior to Claude Code, I have been using Cursor for an year or so.

Sonnet is particularly good at NextJS and Typescript stuff. I also ran this on a medium sized Python codebase and some ML related work too (ranging from langchain to Pytorch lol)

I don't do a lot of prompting, just enough to describe my problem clearly. I try my best to identify the relevant context or direct the model to find it fast.

I made new claude.md files.

Seconded, that a summary description of your problem, codebase, programming dialect in use, should be included whenever a “<Model> didn’t work for me” response.

> But I think we've reached the point where such posts (both positive and negative) are _completely useless_, unless they're accompanied with a careful summary of at least ...

I use Claude many times a day, I ask it and Gemini to generate code most days. Yet I fall into the "I've never included a line of code generated by an LLM in committed code" category. I haven't got a precise answer for why that is so. All I can come up with is the code generated lacks the depth of insight needed to write a succinct, fast, clear solution to the problem someone can easily understand in in 2 years time.

Perhaps the best illustration of this is someone proudly proclaimed to be they committed 25k lines in a week, with the help of AI. In my world, this sounds like they are claiming they have a way of turning the sea into ginger beer. Gaining the depth of knowledge required to change 25k lines of well written code would take me more than a week of reading. Writing that much in a week is a fantasy. So I asked them to show me the diff.

To my surprise, a quick scan of the diff revealed what the change did. It took me about 15 minutes to understand most of it. That's the good news.

The bad news it that 25k lines added 6 fields to a database. 2/3's were unit tests, perhaps 2/3's of the remainder was comments (maybe more). The comments were glorious in their length and precision, littered with ASCII art tables showing many rows in the table.

Comments in particular are a delicate art. They are rarely maintained, so they can bit rot in downright misleading babble after a few changes. But the insight they provide into what author was thinking, and in particular the invariants he had in mind can save hours of divining it from the code. Ideally they concisely explain only the obscure bits you can't easily see from the code itself. Anything more becomes technical debt.

Quoting Woodrow Wilson on the amount of time he spent preparing speeches:

    “That depends on the length of the speech,” answered the President. “If it is a ten-minute speech it takes me all of two weeks to prepare it; if it is a half-hour speech it takes me a week; if I can talk as long as I want to it requires no preparation at all. I am ready now.”

I used PayPal's KYC web interface the other day. It looked beautiful, completely inline with the rest of PayPal's styling. But sadly I could not complete it because of bugs. The server refused to accept one page, it just returned to the same page with no error messages. No biggie, I phoned support (several times, because they also could not get past the same bug), and after 4 hours on the phone the job was done. I'm sure the bug will be fixed a new contractor. He spend an few hours on it, getting an LLM to write a new version, throwing the old code away, just as his predecessor did. He will say the LLM provided a huge productivity boost, and PayPal will be happy because he cost them so little. It will be the ideal application for an LLM - got the job done quickly, and no one will read the code again.

I later discovered there was a link on the page that allowed me to skip past the problematic page, so I could at least enter the rest of the information. It was in a thing that looked confusingly like a "menu bar" on the left, although there was no visual hit any of the items in the menu were clickable. I clicked on most of them anyway, but they did nothing. While on hold for phone support, I started reading the HTML and found one was a link. It was a bit embarrassing to admit to the help person I hadn't clicked that one. It sped the process up somewhat. As I said, the page did look very nice to the eye, probably partially because of the lack of clutter created by visual hints on what was clickable.

[0] https://quoteinvestigator.com/2012/04/28/shorter-letter/

There are some tasks that it can fail and not, but a lot of "Claude Code [or whatever] is amazing" with a reply along the lines of "It doesn't work for me, it just creates a bunch of slop in my codebase." IMO is "i know how to use it" vs "I don't know how to use it" with a side of "I have good test coverage" vs "tests?"

> You can tell Claude to verify its work

Definitely - with ONE pretty big callout. This only works when a clear and quantifiable rubric for verification can be expressed. Case in point, I put Claude Code to work on a simple react website that needed a "Refresh button" and walked away. When I came back, the button was there, and it had used a combination of MCP playwright + screenshots to roughly verify it was working.

The problem was that it decided to "draw" a circular arrow refresh icon and the arrow at the end of the semicircle was facing towards the circle centroid. Anyone (even a layman) would take one look at it and realize it looked ridiculous, but Claude couldn't tell even when I took the time to manually paste a screenshot asking if it saw any issues.

While it would also be unreasonable to expect a junior engineer to hand-write the coordinates for a refresh icon in SVG - they would never even attempt to do that in the first place realizing it would be far simpler to find one from Lucide, Font Awesome, emojis, etc.

> Clear instructions go a long way, asking it to review work, asking it to debug problems, etc. definitely helps.

That's exactly what I learned. In the early 2000's, from three expensive failed development outsourcing projects.

I second this and would add that you really need an automated way to do it. For coding, automated test suites go a long way toward catching boneheaded edits. It will understand the error messages from the failed tests and fix the mistakes more or less by itself.

But for other tasks like generating reports, I ask it to write little tools to reformat data with a schema definition, perform calculations, or do other things that are fairly easy to then double-check with tests that produce errors that it can work with. Having it "do math in its head" is just begging for disaster. But, it can easily write a tool to do it correctly.

Yeah I got about half way through before thinking "wow theres no information in this" and giving up.

I like them for refactoring and “explain this massive codebase please”. Basically polishing or investigating things that already work.

But I think we should expect the scope of LLM work to improve rapidly in the next few years.

https://metr.org/blog/2025-03-19-measuring-ai-ability-to-com...

> They're amazing. But the moment you try to do the more "serious" work with them, it falls apart rapidly.

Sorry, but this is just not true.

I'm using agents with a totally idiosyncratic code base of Haskell + Bazel + Flutter. It's a stack that is so quirky and niche that even Google hasn't been able to make it work well despite all their developer talent and years of SWEs pushing for things like Haskell support internally.

With agents I'm easily 100x more productive than I would be otherwise.

I'm just starting on a C++ project, but I've already done at least 2 weeks worth of work in under a day.

The pipeline dries up when orgs would rather get the upfront savings of gen AI productivity gains versus invest in talent development.

The contractors working on my house sometimes paint a room bright pink for no particular reason.

When I point that out, they profusely apologize and say that of course the walls must be white and wonder why they even got the idea of making them pink in the first place.

Odd, but nice fellows otherwise. It feels like they’re 10x more productive than other contractors.

FWIW I went Claude Max after Pro, and the trick is to turn off Opus. If you do that you can pretty much use Sonnet all working day in a normal session. I don't personally find Opus that useful, and it burns through quota at 5x the speed of Sonnet.