Building a full-stack app with Wasp, an agent-friendly web framework(wasp.sh)
wasp.sh
Building a full-stack app with Wasp, an agent-friendly web framework
https://wasp.sh/blog/2026/03/29/hireveld-from-10-stacks-to-production-with-wasp
3 comments
Yes, the main thing is that we're trying to see if AI can have an easier time using one framework vs. another, and how important it is.
All Rails-like frameworks (Django, Laravel, or Wasp in this case) claim its helpful to use something opinionated and structured, which makes sense (the tradeoff is the flexibility, of course).
We've run some early tests[1] but plan to do a more substantial benchmark next.
[1] https://wasp.sh/blog/2026/03/26/nextjs-vs-wasp-40-percent-le...
All Rails-like frameworks (Django, Laravel, or Wasp in this case) claim its helpful to use something opinionated and structured, which makes sense (the tradeoff is the flexibility, of course).
We've run some early tests[1] but plan to do a more substantial benchmark next.
[1] https://wasp.sh/blog/2026/03/26/nextjs-vs-wasp-40-percent-le...
Django isn't really that opinionated though. It's a common misconception. It's quite decoupled with only the settings file being actually required.
Heh. The problem with Django is certainly not that it has accumulated too much.
Anyway, this seems a bit silly. There's nothing here that is "agentic", and javascript is certainly not a language that is especially suited for LLMs except that the training data is there, but that's even more true of React or Django.