02 / PROTOCOL

Consensus with accountable rules.

A proof-of-stake ledger whose consensus, data layout and execution are explicitly designed around post-quantum authentication.

CONSENSUS

How validators reach agreement.

The paper evaluates consensus families in the absence of practical post-quantum signature aggregation.

01

Tendermint-family BFT

A bonded validator set proposes, prevotes and precommits blocks. Finality depends on weighted quorum certificates.

02

Deterministic proposer rotation

Weighted round-robin proposer selection avoids adding a post-quantum verifiable random function to the initial design.

03

Accountable validator behavior

Signed consensus messages support equivocation evidence and proposed accountability mechanisms.

04

No assumed aggregation

The design does not rely on practical aggregation of ML-DSA signatures; bandwidth is explicitly modeled.

STATE & EXECUTION

A focused core. Clear boundaries.

The first release favors auditable native modules over a general-purpose virtual machine.

01

Authenticated state

Sparse Merkle authentication supports state proofs, alongside block storage, snapshots and indices.

02

Native execution

Accounts and transfers, staking, governance, suite registry, DID registry, credential registry and verification market.

03

No general-purpose VM at launch

VM selection is deferred to a post-mainnet research item, OP11, rather than promised for initial release.

04

Normative specification

The draft describes encodings, domain separation, validity rules and state transitions for interoperable implementations.

DATA LIFECYCLE

Keep the record. Prune the weight.

Separating authorization witnesses from canonical transaction bodies is central to the storage strategy.

01 / SubmitTransactions carry ML-DSA signatures and must satisfy validity and authorization rules.
02 / AgreeValidators reach BFT consensus and form signed commit evidence.
03 / FinalizeFinalized transaction bodies and authenticated state remain available according to protocol retention rules.
04 / PruneEligible signature witnesses can be removed from long-term history, subject to the specification’s safeguards.

The ~95% history-growth figure is an analytical model in the paper, not an observed network benchmark.