by frozenlettuce
0 subcomment
- If you take your idea and add some web ui to manage the creation of apps and connect the pieces, I believe that there's great potential!
A couple years back I had a similar idea, but with a postgres+deno, and using .md files as the spec that generated the code https://github.com/lfarroco/verbo-lang
I think that in the future we might have specialized agents that operate under a specific opinionated tech stack. Like having one that is specialized in creating cli apps, another one for react+django+whatever, etc.
by koakuma-chan
2 subcomments
- You eliminate the schema management problem by not having a schema at all? Also, what do you mean schema management problem? I have never had an issue with that when using LLMs.
by fd-codier
1 subcomments
- Maintaining a codebase with mongodb db is already hard enough considering 99% of the time you need a relational db. It always end up as a mess.
But letting an llm doing this as well, ouch. I fear for the maintainability of the codebase in the long term.
by joshribakoff
1 subcomments
- You use a static typed language for guardrails but then you throw out the guardrails of a database schema? Seems like those two decisions are directly at odds.
Without a db schema, you still have to worry about migrating data at runtime or otherwise. Removing the schema just shifts the pain doesn’t remove it, in my experience.
by jpartridge
1 subcomments
- I like this take on the AI app builder hype. Most tools focus on generating an initial UI but are no help with 'seams' like auth and database migrations. Using MongoDB’s flexible schema as the backbone is a smart move for agents that usually hallucinate SQL relations. Will cool to see how the built-in observability handles production incidents.
- When an experienced team with production users first feels ‘we can’t keep duct-taping this,’ what exact failure makes them reach for Modelence instead of just adding another managed service or framework?
by Soerensen
1 subcomments
- The TypeScript + MongoDB combination for AI coding is a smart architectural choice. I've found that schema-less databases reduce the class of errors agents struggle with most - the migration/schema drift issues that require understanding of state over time.
Question: How are you handling the built-in auth when users want to extend it? For example, adding OAuth providers that aren't pre-configured, or custom claims/roles logic. Is this something the framework supports as extension points, or would users need to fork/modify core auth code?
The Claude Agent SDK integration is interesting - have you found specific prompting patterns that work better for TypeScript generation vs other languages? Curious if the type system actually helps agents self-correct as expected.
by gr4vityWall
1 subcomments
- How does your framework compares to Meteor.js? I see similarities in the problems being solved, and the tech stack being used. Do you have examples of the idiomatic way of client/server communication in Modelence?
I think the line between the framework and the AI code generation tool is blurry.
by stoicsession
1 subcomments
- Have you checked out MainMVP (https://www.mainmvp.com) yet? It is completely stack-agnostic and gives you complete freedom in choosing it. You can also use your own keys (BYOK).
by jellyotsiro
0 subcomment
- sick
- To confirm I have the option to run essentially the whole stack on my own in my own hardware if I so choose? That's the ultimate unlock of Open Source trust and no lock-in. Makes it much easier for me to adopt you if you have that
by chaostheory
3 subcomments
- Not keen on your data store choice. Mongodb introduces a lot of other problems. The problem with schemas is an easier one to solve imo
Hacker News