Slice: Define canonical observer coordinate digest domains

[flyingrobots]

Slice

Specify canonical commitment/digest domains for observer-coordinate evidence before adding proof systems.

Done when:

• coordinate, basis/frontier, aperture, posture, reading envelope, retained evidence, and proof references have domain-separated digest labels;
• digest inputs are canonical and versioned;
• profile participants can cite these domains without knowing another participant's internals;
• downstream apps can display/carry the digests as evidence references.

Parent: #58.