Claude Sonnet 4.5

HAHAHA - I like me. but claude (sonnet 4.6) seemed like it was cheerleading a bit

This was actually entirely vibe coded by Sonnet 4.6, with a lot of me yelling at it! It's essentially a SolidStart SPA with virtualized scrolling and a few other tricks (even I don't know most of them!). Vector search is entirely client-side with transformers.js and CLIP. The first load is quite slow unfortunately, cause it has to download the index of photo id -> link (~7 mb last time i checked), and same for searches, as it has to download the vector index (~28mb) and embedding model the first time. Caching is pretty good though. I was previously using [Lychee](https://lychee.electerious.com/), which worked well but wasn't really suited for this use case. So I scrapped it and started from first principles with Claude. This entire task only used up 50% of my 5-hour quota on the $20 plan! Since the site itself is an SPA and images are static, using a VPS is overcomplicating things since I (well, Claude) am essentially just using Caddy as a glorified CDN. But I had free DigitalOcean credits since I'm a student, and where else can I host 40 gigabytes of photos for free? All in all, this was definitely a very fun exploration of what someone who's technical[1] can do with Claude. The code is all open-source (but it's slop) at https://github.com/aadishv/catapp. [1]: at least I'd like to consider myself technical!

I like writing new, interesting code, but learning framework #400 with all its own idiosyncrasies has gotten really old. I just rebuilt a fairly simple personal app that I've been maintaining for my family for nearly 30 years, and had a blast doing with an AI agent - I mostly used Claude Sonnet 4.5. I've been dreading this rebuild mostly because it's so boring; this is an app I built originally when I was 17, and I'm 43 now. I treated Claude basically like I'd treat my 17-year-old self, and I've added a bunch of features that I could never be assed to do before.

I don't get how these agents can work when even Claude Sonnet 4.5 (for example) needs a lot of hand-holding for basic, simple bugfixing stuff. Wouldn't the agents just be huffing and puffing their way off the rails all the time?