Base Facts and Proof Bundles
The smallest useful building block is a fact bundle with proofs.
Base Facts
A base fact is a peer-authored claim such as:
delivered(milestone M, artifact A)accepted(order O)revoked(credential C)
It is authored by one peer and committed into that peer's tree.
Proof Bundle
A proof bundle typically carries:
- canonical facts
- inclusion or exclusion proofs
- referenced peer roots or checkpoints
- peer identities or key references
- enough metadata to interpret the facts
Design Goal
A verifier should be able to:
- recompute the relevant fact commitments
- verify the proofs against trusted roots
- evaluate the protocol rule
without needing the full peer log.
Distribution Pattern
Proof bundles should usually be requested, not broadcast.
The public anchor lets a verifier find the relevant on-chain root state. The proof bundle is then pulled from the holder for a specific claim or protocol transition.
This avoids pushing every possible proof to every possible verifier, and it keeps disclosure scoped to the policy being evaluated.
Common Mistake
Do not confuse:
- the fact bundle
- the proof bundle
- the derived or certified claim
They are different layers and should be modeled separately.