Create a passkey-crypto library for when dealing with more cryptographic algorithms than p256. Currently that is the only algorithm implemented so it is assumed in the methods. But seeing as we need a certain amount of crypto agility and that;
- NIST has started to no-longer recommend
p256 in favor of p384 or p512 and
- We'll need to start working towards PQ signature algorithms
- Support for other crypto library backends for other consumers of this library.
We need a way to reliably go from a CoseKey to a private key and back. Also having helper methods to create a public CoseKey directly from the private key would be a nice simplification in the authenticator code.
This could also be the place where a replacement implementation for the coset::CoseKey type and other types we use from coset go so that we can yank it from our tree. That library has been a source of frustration on multiple occasions.
Create a
passkey-cryptolibrary for when dealing with more cryptographic algorithms thanp256. Currently that is the only algorithm implemented so it is assumed in the methods. But seeing as we need a certain amount of crypto agility and that;p256in favor ofp384orp512andWe need a way to reliably go from a
CoseKeyto a private key and back. Also having helper methods to create a publicCoseKeydirectly from the private key would be a nice simplification in the authenticator code.This could also be the place where a replacement implementation for the
coset::CoseKeytype and other types we use fromcosetgo so that we can yank it from our tree. That library has been a source of frustration on multiple occasions.