I played around with Claude Code today, first time I've ever really been impressed by AI for coding.

Tasked it with two different things, refactoring a huge function of around ~400 lines and creating some unit tests split into different files. The refactor was done flawlessly. The unit tests almost, only missed some imports.

All I did was open it in the root of my project and prompt it with the function names. It's a large monolithic solution with a lot of subprojects. It found the functions I was talking about without me having to clarify anything. Cost was about $2.

Yes, massively.

There’s a learning curve to it, but it’s worth literally every penny I spend on API calls.

At worst, I’m no faster. At best, it’s easily a 10x improvement.

For me, one of the biggest benefits is talking about coding in natural language. It lowers my mental low and keeps me in a mental space where I’m more easily able to communicate with stakeholders holders.

Really great for quickly building features but you have to be careful about how much context you provide i.e. spoonfeed it exactly the methods, classes, files it needs to do whatever you're asking for (especially in a large codebase). And when it seems to get confused, reset history to free up the context window.

That being said there are definite areas where it shines (cookie cutter UI) and places where it struggles. It's really good at one-shotting React components and Flutter widgets but it tends to struggle with complicated business logic like sync engines. More straightforward backend stuff like CRUD endpoints are definitely doable.

It gets confused if I give it 3 files - how is it going to scan a whole codebase and disparate systems and make correct changes.

Pah! Don't believe the hype.

Yes, it helps me write SQL queries in seconds that I otherwise would spend days on or give up completely.