📝 Nuxt Content Assets and Nuxt 4 / Nuxt Content 3

[davestewart]

Hi everyone,

First of all my sincere apologies for ignoring this package this year 🙏

As I mentioned in various other posts I've just not had the bandwidth to be in the OSS game – nor Nuxt Content Assets which itself is a pretty intense set of workarounds to achieve its transparent DX! Also, I made various promises to get back on this which I never delivered on (and that's been weighing heavy on on my mind!) so I'm sorry for raising hopes or if it affected anyone's timelines.

Anyway, the time has now come for me to migrate to Nuxt 4 and Nuxt Content 3 so I've made a project outlining all the outstanding work (migrations, missing features, outstanding bugs):

• https://github.com/users/davestewart/projects/3/views/1

I am actually working on it right now, so my head is back in the Nitro / Nuxt / Nuxt Content / Nuxt Content Assets code and I'm feeling good about it. In the ideal world, this will all be done by Christmas end of Jan 2026.

Watch this space! ⬆️

[Qhbee]

Hi @davestewart,

First of all, thank you for developing and maintaining nuxt-content-assets. It is an incredibly valuable module, and I heavily rely on it for my projects.

I noticed that this issue regarding Nuxt 4 and Nuxt Content 3 compatibility has been pending for a while. Given how crucial this upgrade is, I would like to fork the repository and try to implement a fix to help move this forward.

However, I want to be completely transparent: my current expertise in the underlying architecture might not be sufficient to solve this entirely on my own, so I plan to rely on AI tools to assist with generating the code.

Before I start, I want to respectfully ask: Do you accept PRs that contain AI-generated code?
Please rest assured that I will not submit raw, unchecked outputs. I will thoroughly review, test, and debug the code to the best of my ability to ensure it aligns with your project.

If you are open to this approach, please let me know if you have any specific architectural guidelines, testing requirements, or other expectations for this PR. If not, I completely understand and respect your repository's policies.

Thank you for your time and your continuous work!

[davestewart]

Hey @Qhbee – thanks for your kind words and your enthusiasm!

You're right, I haven't managed to find the time to update the repo to Nuxt 4 and Content 3.

I would like to have a go myself at making it work, as I do have a firm approach in my head on how I would like to do it. And to tell you the truth, I will probably also use AI to help me!

But I would be interested to see what an AI could do by itself, so if you want to have a go, go for it. The main change from V2 to V3 is that the transformed content is now organised "by collection" and stored in a database. Honestly the query API seems like it's a significant change as well. I set up one project and everything seems much more defined, especially with the collection types:

https://content.nuxt.com/docs/getting-started/migration

So TL;DR, have a go. If I do it at the same time, I may not accept yours, but if I don't, I probably will. Either way, it will be an interesting experiment.

[Qhbee]

Thanks for the detailed reply and the green light, @davestewart!

I completely understand your position. I will treat this as a great learning opportunity and a fun AI experiment. If you end up implementing your own vision in the meantime, I'd be more than happy to use your official version. No pressure at all regarding merging my future PR.

I will carefully review the migration docs you provided, focusing on the new "collections" structure, the database approach, and the updated Query API.

I'll start experimenting in my fork and let you know if I make any meaningful progress. Thanks again!

[davestewart]

Nothing like a bit of healthy competition!

[Qhbee]

Hi again @davestewart,

As I am quite new to open-source contributions, I am not entirely familiar with the standard conventions for major version updates. Therefore, I have just one quick architectural question before I dive too deep into the codebase:

Should this PR aim for backward compatibility with Nuxt Content v2, or should it be a clean break that strictly targets Nuxt Content v3?

Given the fundamental differences (files vs. collections/database), supporting both simultaneously might complicate the module significantly. I assume targeting v3 exclusively might be the preferred route (perhaps for a new major release of nuxt-content-assets), but I want to align with your roadmap.