gm @TxFlow_L1 is interesting for one reason people often gloss over: proving a tx is valid ≠ deciding where it belongs. That gap is where things get spicy. A leader-based BFT can look great on paper—until a liquidation storm hits and the system suddenly needs the most throughput. That’s the stress test I’d care about. How does TxBFT resolve concurrent DAG batches with no causal relationship into a deterministic total order? And is that pipeline actually running end-to-end today, or does the single sequencer still sit in the critical path? The real test: take it outside the trusted/core operator set and hammer it with load. Because DAGs don’t prove much until multiple independent operators are actually feeding them. @matthubuilds @kaisrr3a