With Buildernet under active development, it's timely to review other approaches to encrypted mempools. This study considers both mempools that support programmability and those that aim to fully prevent order flow manipulation until inclusion in a block.
Notably, projects like Osmosis, which initially advocated for fully encrypted mempools, have begun relaxing that stance, gradually permitting specific forms of MEV extraction (e.g. the ProtoRev module – see appendix to FRP). Meanwhile, recent advances such as Ferveo and iterations on Batched Threshold Encryption (Usenix24, BEAT-MEV, and the Usenix24 follow-up) have begun addressing previous bottlenecks in decryption throughput.
Related Projects
This line of work aligns well with the Encrypted backruns direction, which can be seen as an initial step toward a programmable encrypted mempool designed for backrunning. As more foundational primitives are identified, we intend to publish accessible write-ups and release corresponding prototypes to support further development of MEV-aware encrypted mempools.
This effort also overlaps with our other efforts aimed at Removing Relays in the Block Building Pipeline, given their shared reliance on threshold cryptography and similar design constraints.
Questions
- How can state-of-the-art threshold encrypted mempools be adapted to support programmability?
- What is the minimal set of features required to enable MEV extraction in an encrypted mempool?
- What are the implications of eliminating relays in a network with encrypted mempools?
