Description of your request or bug report: :warning: tl;dr use AI to suggest tags based on the blurb and/or pull more info from sources I think most content on Natively is lacking genre/content tags, which makes it difficult for users to find media they’d be interested in. While the level does pla…

This seems a bit related to: [image] Automated tag handling/prompting for new entries Product Requests Description of your request or bug report: Following discussion here: Tags, Tags, Tags [Official Management] - #316 by 暁のルナ . What I said there (with some ad…

This has just reminded me that another option might be to use AL’s tags for anime, manga, and light novels, as the majority of their db is well-tagged.

I heavily sympathize on the “new content needs tags” thing, I’m skeptical of an AI-based approach [image] bibliothecary: This has just reminded me that another option might be to use AL’s tags for anime, manga, and light novels, as the majority of their db is well-tagged. This seems like a b…

[image] 暁のルナ: I’m skeptical of an AI-based approach This is why I left them separated, I have a bit of skepticism as well.

Cool. I just meant to reply with that & elaborate back in February :sweat_smile: … So figured I’d chime in

[image] 暁のルナ: I just meant to reply with that & elaborate back in February :sweat_smile: … So figured I’d chime in ▶ Offtopic-ish

Just to point out the AI is just one suggestion, the other is to pull more info directly from the item page categories/tags on Amazon/Yes24/Ridibooks. :slightly_smiling_face: [image] 暁のルナ: This seems like a better potential approach imo. Assuming AL is Anilist? Weren’t a lot of tags here popula…

[image] bibliothecary: @Megumin The link you added to the OP is for this thread :sweat_smile: It’s clear that I can’t be doing too many of these at the same time, I start making mistakes :sweat_smile: Fixed, thanks for pointing it out

Suggestion: Provisional tags based on blurb

Natively Product Requests

Megumin March 24, 2026, 8:52am 2

This seems a bit related to:

I’ll leave them as separate entities for now because of how different the approach is between the two threads, but they can probably be handled together as they seek a similar outcome.