RE: LeoThread 2026-04-19 22-29

[eatthebite]

Migrating BFT from Tendermint to Substrate would address this. That would require rewriting the stack or forking Serai as a base. Additional features to include:

Comments

[eatthebite]

• On-chain privacy: encrypt protocol transactions and use zk proofs for consensus so transaction details remain private
• Delegated proof-of-stake: enable token holders to stake with any chosen node

[eatthebite]

• Mempool operations: support limit orders plus cancel-limit so a limit can be cancelled before processing
• Flash loans: allow borrowing from the protocol with repayment within the same block

[eatthebite]

• Replace the v2 AMM with intent-based matching, turning assets into lending collateral for flash loans

Layering an advanced swap queue on top of this design would outperform HyperLiquid and NEAR, but it hasn’t been done yet

[eatthebite]

$XMR is proving difficult to meet THORChain integration standards
Adding LPs and swaps through a single vault was completed a week ago without issues

[eatthebite]

However, the protocol uses multiple vaults, so some nodes lack a complete vault view. Transaction view keys (to audit outbounds) must be propagated
Propagation implemented. Vault churn and migration require multi-vault scanning

[eatthebite]

Implemented. Old vaults must be monitored for refunds, so old vaults need to remain available on active and standby nodes, requiring view-key scanning for all vaults

[eatthebite]

Implemented. Vault solvency monitoring requires key-image generation and propagation across all vaults and nodes
Implemented. For fault tolerance, if a node goes offline it must repair and rebuild the full XMR vault inventory

[eatthebite]

This is the current stage