Ahoy 
ETH Global hackathons are always interesting. They often attract a mix of new developers who have barely touched web3 before, alongside some very experienced pirates who have been around since the start. ETH NY saw this same pattern.
Open/Innovation Track - $1500 - Unveiling the MEV Potential
We awarded this prize to a team called “Integrity Proof”, who developed a way to ensure the integrity of bundles using ZKP. Proof of Bundle Integrity allows bundle senders to specify the exact slot they wants their bundle to be placed in a block and trustlessly prove whether a block builder has tampered with the bundle, leveraging zero knowledge proofs.
This goes a long way to solving the trust issue between block builder and bundle sender.
Ecosystem Track - $2000 - Broaden the MEV Landscape
We awarded this prize to a team called Lime, who developed a way to handle pricing offchain in a manner that we thought could be applied in an interesting way on SUAVE.
Lime is an active hook manager which allows fillers or market makers to set prices and fill Intents / RFQ based swap requests. This system is intended to be used through UniswapX or other intent based mechanism. Lime believes that RFQ/Intents are the future of MEV.
Data Analysis Track - $1500 - Illuminate the MEV Supply chain
We awarded this prize to a team called MEVictim Rebate. While this is not specifically a data project, we felt that their use of a subgraph in order to collect information about MEV used to allocate NFTs, which can then be used on other domains for rebates came closest to the goal of this particular track.
Their aim is to design better market incentives by identifying MEV victims using historical on chain data and then airdrop victims a token to access a token gated Uniswap v4 pool on Scroll. They aim to showcase the detrimental impact of MEV attacks on users and highlights how it is possible to detect MEV attacks using Axiom to qualify wallet addresses for benefits in the next generation of DeFi protocols.
General Thoughts
We received 8 submissions in total, which is fairly high given that we did no advertising and were one of the minor sponsors. Major sponsors, with bigger prizes and greater presence, tend to receive 30-40 submissions. Receiving 8 is indicative of a growing interest in MEV and the tools Flashbots has made available. As we move towards SUAVE, and the need to cultivate a network of people with different skills, backgrounds, perspectives and experience levels, these kind of events will likely become even more important.
While some were elementary, they nevertheless illustrate interesting and novel ways people new to MEV conceptualise it, which can help us iterate over and improve our products. In this sense, such hackathons are great venues for in-the-wild UXR. They also serve as wonderful places to nerdsnipe up and coming talent: something we already do fairly well.
On that note, some interesting students who run the University of Purdue Blockchain Club presented sndwch_protocol (code, paper), which proposes a (fairly naive) method to mitigate Sandwich Attacks in DeFi using Uniswap V4’s action hooks and the OZ defender. We talked at length about their approach and they will continue to work on it and iterate towards something better over time. I personally think that teams like this are good to focus on and develop: they’ve won previous hackathons, but this time wanted to “do something long-term and meaningful”.
In general, their attitude highlights the incentive landscape at these events. Most teams try and do generic projects which use as many sponsor’s tech stacks as possible in order to make them eligible for the most prizes. This often comes at the cost of doing something really great, focused and meaningful. Nevertheless, over time, people fall into the rabbit hole and - after a few experiences - tend to start optimizing for work that will get them noticed by more skillful, long-term focussed people.
Other Submissions
- Solvify. A solver implementation for an intent-centric future on Etheruem - all with a delightful web2-like UX and powerful AI agents from LangChain to optimally satisfy intents and retain value for users.
- Arb Controller. A uniswap v4 hook that sets dynamic fee for a pool based on the price movements. The dynamic fee partially discriminates informed order flow from arbitrageurs. It gives LPs similar efficiency to the dynamic spreads in tradfi, while still allowing LPs to be passive.
- SUAVE intents. An exploration on using Suave as a privacy preserving intent mempool.
- Oracle Flow. A more efficient solution for oracle updates using MEV-Share.
I look forward to many more such events and all the cool ideas that will come from them, especially as SUAVE begins to get used in the wild.