0
Resolved 17/Jun — was the most severe gap of all. Every SNTI block ever mined violated BIP141's coinbase witness rule, meaning no node other than the one that mined a block could ever accept it. This was undetectable with a single node and was only found once a second independent node was actually stood up. Fixed via a stable preimage signing design — see Changelog Phase 6.9. This is the strongest argument yet for never skipping the multi-node test, no matter how much single-node testing looks clean.
1
No external eyes have looked at this code. Every fix so far has been found by this team, in this session. A second reviewer — even an informal one — would likely find things missed here.
2
No adversarial testing. Nobody has tried to break the PoUW validation, attempt a key-reuse attack, or fuzz the XMSS bridge.
3
No multi-node network has ever existed. All consensus behavior observed so far is single-node. Network partition handling, peer banning, and propagation under real latency are unverified.
4
Mainnet difficulty has never actually been mined. The 0x1d00ffff target is copied from Bitcoin's genesis difficulty but untested against this codebase's actual hash rate plus PoUW overhead.
5
No incident has happened yet. There is no tested procedure for what happens if a bug is found in an already-mined block — no chain-split contingency, no communicated rollback plan.