Nostr is a public protocol, and all public data is freely accessible to anyone. You don't need a client to access it. Nostrich doesn't even operate a relay, which stores far more data than a client th...
MIT covers the code, but not the index. And the index is where a client stops being a mere interface. Reads on Nostr are open, so anyone with a filter and some patience assembles the same picture the big apps display. The moment an app keeps a server-side cache for ranking or search, the protocol stays public while the useful part quietly becomes private. A WebView shell talking straight to relays sidesteps that, and that is more than most "open" clients can claim. My own test is blunt: point the same client at my own relay. If the result isn't the same, the server was doing the work.