Testnet-2 has been running since 24 October 2026 with the leader VRF moved to RFC 9381. Mainnet has not launched. Testnet coins have no value. Current status and one-command node setup are in the repo and on the project site.
I am looking for independent critique of the core claims before mainnet, particularly:
-
**At most one transition finalizes per height** under the stated Byzantine and network assumptions (f < n/3, asynchronous network). Spec: `spec/TWO_PHASE_FINALITY.md`. Implementation: `src/node/node_consensus.cpp` and `src/consensus/canonical_derivation.*`.
-
**Confidential transactions cannot create value.** Pedersen commitments + range proofs (secp256k1-zkp). Balance is checked homomorphically; excess authorization and non-negative range proofs are required. Spec: `spec/CONFIDENTIAL_UTXO_SPEC.md` §§3, 4, 7, 15. Code: `src/crypto/confidential.*` and `src/utxo/validate.*`.
-
**Deterministic derivation of committee, checkpoints, and rewards** from finalized history only. No speculative views. Specs: `CHECKPOINT_DERIVATION_SPEC.md`, `COMMITTEE-SELECTION.md`, `ECONOMICS.md`.
The full list of claims, reading order, and how to test them is in `docs/REVIEWING.md`. Findings can still change the protocol. Report security issues privately per `SECURITY.md`; other feedback can go here or as GitHub issues.
Repo: GitHub - FinalisCore/finalis-core: BFT blockchain with instant finality, confidential transactions from genesis, and a trust-minimized light-client wallet. C++20. · GitHub
Whitepaper and specs are in the repo. I am also looking for additional independent validators on testnet-2.
Questions or counter-examples are welcome.