SecureEnclave.PrivateKey properties

Hi,

Is there some reference documentation about the properties of a CryptoKit SecureEnclave PrivateKey and its properties?

Concretely, these are some of the questions that I wanted to find a (documented) answer on:

• Who can use a SecureEnclave.P256.*.PrivateKey if they have access to the dataRepresentation? I expect that the private key is bound to the specific secure enclave processor, but it also seems to be bound for the user that created the key (from observation by creating a PrivateKey without any access control). What if there's a restore from backup of the machine, will the private key still be usable?
• What does a SecureEnclave.P256.*.PrivateKey's dataRepresentation include? From observation, I'm assuming the dataRepresentation is a signed/encrypted blob that includes a unique ID (no 2 keys are the same), the access control settings (biometry required, passcode required, ...), some sort of version of the biometry (so it is be invalidated when the biometry changes). Is there anything else? I'm not interested in the actual encoding (which I understand is undocumented), but want to get an idea of what properties are included in the representation and e.g. can't change in the future.

Answers to these questions could e.g. help make a decision how secure the private key's dataRepresentation needs to be kept (e.g. if it can only be used by myself, and i'm sure it will only ever be valid with the access control flags its representation contains, I could decide it's ok to have this key be in a public place)

I tried looking for answers in some pieces of documentation, but couldn't immediately find the details I was looking for:

• The CryptoKit SecureEnclave documentation
• The Secure Enclave article
• The Protecting keys with the Secure Enclave article

thanks!

Remko