More a discussion then an issue really.

We are using an independent implementation of ATProto cryptography that uses ring for the actual cryptography part of p256. Is there a general interest on having that ported back to atrium-rs? The implementation look quite similar so I guess the effort would not be too much.

Aside from that we plan to support different providers (such as Security.framework and Web Crypto API) for our implementation. Since this will involve quite some changes on our end we might as well integrate this into atrium. This will require some considerable changes though - is this considered out of scope or potentially interesting?

If out of scope, we might provide this as an external crate.