MEV Boost Account Abstraction Based on ERC-4337

Project Abstract

Only validators and searchers benefit from current MEV ecosystem while the transaction senders where MEV is actually extracted from are excluded.

In fact, senders need to pay twice for the transaction execution: the transaction fee and the fetched MEV. The latter always goes against the intention of the sender and can be consider as a robbery in secret towards sender.

The project supports senders to capture MEV under the ERC-4337 framework:

• Senders use special ERC-4337 wallets called MEVBoostAccount and send special userOps called BoostUserOp which invokes boost methods of MEVBoostAccount.
• Searchers use a special ERC-4337 paymaster called MEVBoostPaymaster which refunds maxium MEV to users by BoostUserOp auction and also pays transaction fee for BoostUserOp.

The project mainly has following advantages:

• Improve the current MEV ecosystem and maximize the sender's interests by trustlessly ensuring the maxium MEV capture by the sender.
• It provides a completely decentralized and permisionless MEV-Share solution.
• Considering the existence of MEV, the project promotes the adoption of the ERC-4337 wallet by providing the build-in mev capture features and making transaction possibly less expensive than the EOA.

Objectives

• Implement the MEVBoostAccount and MEVBoostPaymaster that conform to the ERC-4337 specification.
  • The project will be open source and provide smart contracts and sdks to support third party integration.
  • After wallet sdk integration, users can easily deploy MEVBoostAccount and send BoostUserOp through the sdk.
  • After seacher sdk integration, searcher can easily fetch BoostUserOp through the sdk and provide MEV to the BoostUserOp.
• Further implement the BoostUserOp auction mechanism and provide various auction modes.
  • The project will be open source and provide smart contracts and sdks to support third party integration.
  • Wallet users are guaranteed to capture as much MEV as possible through the auction mechanism.
  • Wallet users can choose the appropriate auction mode for each BoostUserOp through the sdk.
  • Searcher can obtain latest MEV quotation of BoostUserOp and bid on demand through sdk.
• Since MEV is ubiquitous on Ethereum, implement a modified EntryPoint which natively supports BoostUserOp.
  • Reduce gas consumption of BoostUserOp.
  • Simplify the development of MEVBoostAccount and MEVBoostPaymaster.

Outcomes

• Improve the current MEV ecosystem and maximize the sender's interest by trustlessly ensuring the maxium MEV capture by the sender.
• It provides a completely decentralized and permisionless MEV-Share solution.
• Considering the existence of MEV, the project promotes the adoption of the ERC-4337 wallet by providing the build-in mev capture features and making transaction cost possibly less expensive than the EOA.
• The sender's interest are protected for searchers get involved in sufficient competition through the auction mechanism and public mempool.
• Expand the application scenarios of the ERC-4337 framework.

Background

I have recently started the MEVBoostAA project and verified my idea. Here is the project link: MEVBoostAA.

Current MEV Ecosystem

• In the view of the sender
  • sender --[fee to excute tx]--> validator
• The fact goes against the intention of the sender
  • sender --[fee to excute tx]--> validator
  • sender --[mev caputured by searcher]--> searcher
  • searcher --[share mev with validator]--> validator

MEV Ecosystem with MEVBoostAA

MEVBoostAA captures MEV from searchers to senders based on the framework of ERC-4337. so, senders will be protected.

• sender --[mev caputured by searcher]--> searcher
• searcher (as a paymaster) --[fee to excute tx] --> validator
• searcher --[maxium MEV refund]--> sender
• searcher -- share mev with validator --> validator

Some Related Links

• MEV and Me
• MEV-Share: programmably private orderflow to share MEV with users
  • By comparison, this proposal has the following advantages:
    • It is a completely decentralized and permisionless MEV-Share solution.
    • Utilizing the ERC-4337 framework, this proposal is more concise and with less work. This proposal only need to foucus on the development of MEVBoostAccount and MEVBoostPaymaster.
    • This proposal guarantees the interest of senders at the constract level without introducing trust assumption.
    • Easy to integrate without integrating the heavy flashbot service.
    • This proposal is premissless and with high transparency. The BoostUserOp can be public in the ERC-4337 mempool.
• awesome-MEV-resources

Timeline

Note: assume the proposal would be granted from May 2023.

It is expected to take 9 months and the roadmap is as follows:

• May 2023 - July 2023 (3 months)
  • Implement the MEVBoostAccount and MEVBoostPaymaster to supports senders to capture MEV under the ERC-4337 framework.
  • output
    • A solution that supports senders to capture MEV under the ERC-4337 framework
    • Smart contract
    • SDK for sender
    • SDK for searcher
• August 2023 - October 2024 (3 months)
  • Further implement the BoostUserOp auction mechanism and provide various auction modes.
  • output
    • A solution that supports BoostUserOp auction and various auction modes.
    • Smart contract
    • SDK for sender
    • SDK for searcher
• October 2023 - January 2024 (2 months)
  • Implement a modified EntryPoint which may includes the utility functions, interaction interfaces and integrated BoostUserOp handling.
  • output
    • Modified Entrypoint contract optimized for BoostUserOp.
    • A report that compares the complexity of the implement of MEVBoostAccount and MEVBoostPaymaster after modification.
    • A report that compares gas consumption after modification.