I’ve been seeing a lot of projects lately labeled as “Built with Opus 4.7,” especially with tools like Claude Code.
Out of curiosity, I went through a few of them to understand how far this actually goes beyond demos.
At a surface level, it’s honestly impressive:
Full app scaffolding
Working frontend + backend flows
Iterative fixes with minimal input
But when I tried to think about these from a production perspective, a few things felt unclear:
How are people handling architecture decisions long-term?
What happens when the codebase grows beyond initial prompts?
Who owns debugging when things break in production?
It feels like AI is great at getting things running, but I’m not fully convinced about how well it holds up over time.
I wrote a deeper breakdown of what I observed here if anyone’s interested: I Looked at “Built with Opus 4.7” Projects. Here’s What I Think
Would love to hear from people who’ve actually shipped something using this approach. Is this replacing parts of your workflow, or just speeding up prototyping?
No responses yet.