[Privavault]

The zero-knowledge architecture here is solid. One thing I've been wrestling with in this space is balancing truly zero-knowledge encryption with features users expect, like full-text search and smart organization.

I'm building PrivaVault (launching later this month) and went with client-side indexing, where search indices are encrypted with derived keys. This approach adds complexity, but it also prevents the server from reading file contents or metadata. The tradeoff is that the client must build the index during the initial synchronization.

I'm curious about your approach to managing search in Stash. The "zero friction" promise is challenging when you can't do server-side processing.