I think the most important (although I'm biased because I wrote both of them) are NIP-31 and NIP-89.

Both of these combined provide magical interoperability across use cases at O(1) complexity (meaning, adding new use cases to existing applications is basically free instead of having to linearly add each individual feature an application might want to support).

The reason I say this is that currently we are abandoning platform siloes and sleepwalking into use-case/kind siloes.

Meaning, the microblogging nostr apps compete independently against Twitter.
The youtube-like nostr apps compete independently against Youtube/Vimeo.
Etc, etc.

Sure, there is some cross-polination, but perhaps 10% of what it could be if we allowed the discoverability layer to extend to it's full conclusion.

Dethroning network effects is very hard, and we should leverage the properties the protocol extends to make the most out of the symbiotic relationships the different use-cases built on nostr can have.

I wrote about this on https://highlighter.com/pablof7z/38iu6ttnk6zw3wdd
 

 Link doesn’t work - want to read more about this! 

 Works! 

 There isn't nearly enough adoption of either NIP-31 or NIP-89. Incredibly powerful combo for "other stuff" discoverability and showcasing protocol interop.

nostr:nevent1qqsvgv472q48h32ajgghwjqtf5d7m8t5yykjqyuyeg29e0jxsyeuezsppemhxue69uhkummn9ekx7mp0qgs04xzt6ldm9qhs0ctw0t58kf4z57umjzmjg6jywu0seadwtqqc75srqsqqqqqp8gq2gt 

 going to check out the two nips right now Pablito .  

 i agree. the nips can both solve this issue. 
nip31 ALT text to describe event better. . here user doesnt even know what kind {interger} is. 
also, nip89 for discoverability, would be nice if the app recommended a client that opens those kind of events. "unknown" 

what do you think? . hope im not  muted
nostr:note1df0zraglzx9t7psrhsd5pvrw6yc5ygj0zxkhf9tvrz3hgmp3nz3qw0y958