RFP: AI-Agent-Driven MEV and Transaction Ordering Simulator on Gnosis Chain| OLAS x Shutter DAO 0x36

Shutter DAO DAO Proposals
1
Proposal Title Author(s) Phase Type Date Created
RFP: AI-Agent-Driven MEV and Transaction Ordering Simulator on Gnosis Chain brainbot gmbh I On-chain 06.06.2024

Joint Call for Development: OLAS x Shutter DAO 0x36 Collaboration on Gnosis Chain

Proposal

If this proposal passes, Shutter DAO 0x36 in collaboration with OLAS issues a request for proposals (RFP) to the developer community. This RFP aims to attract developers and teams to build an AI-agent-driven simulator for MEV and transaction ordering strategies on Gnosis Chain. The focus is to leverage the shutterized, encrypted mempool and integrate OLAS autonomous agents, establishing a playground for the practical exploration of MEV solutions and transaction ordering strategies.

With the shutterized Gnosis Chain going mainnet within the next couple of weeks, we’re excited to share this collaboration, which proposes a blend of OLAS’s autonomous agents and Shutter’s L2 encrypted mempool to create a dynamic testbed for MEV and transaction ordering strategies.

The idea is that LLM-powered autonomous agents are interacting in a low-stakes, low cost, yet mainnet environment and with real funds. Some of these agents would use the encrypted mempool, some would not. We can then monitor and analyze the economic and technical implications of this behavior and use the feedback from it to improve the system and parameters.

Successful applicants will share rewards up to 50,000 USDC, 500,000 SHU and can expect additional rewards through the OLAS developer revenue share program.

Scope of Work

Set Up of OLAS Autonomous Agents Framework

  • Development of Autonomous Agents: Modify OLAS’ already existing suite of trading agents to create and deploy autonomous agents capable of simulating a range of behaviors within the MEV ecosystem, from normal user trading to front-running strategies.

Creation of OLAS Agents for Simulation

  1. MVP Implementation: a) Simulate normal user trading behavior b) Implement standard MEV/front-running strategies c) Operate (or ensure the operation of) the agents

Sidenote: The agent system could already be tested on the shutterized Chiado testnet (The First Shutterized Testnet Is Now Live on Chiado!) and the Swapr V3 deployment which is live on top of it.

  1. Dashboard:

Develop a simple dashboard which can be used to supervise the state of the autonomous agents, including their individual P&L and the amount of front-running that they engaged in or were a victim of.

  1. Optional Enhancements: a) Liquidity Providing (LPing). b) Minting new tokens.

Encrypted Mempool Testing

  • Performance Comparison: Set up a longer term experiment where, e.g. 10 agents utilize the Shutter encrypted mempool and 10 do not, to observe and compare their trading performance.

Once the more technical aspect of this experiments are set up, a follow-up proposal could be issued to Shutter DAO 0x36 to fund and further incentivize the autonomous trading agents and liquidity.

Rewards and Incentives

Successful applicants will share rewards up to 50,000 USDC, 500,000 SHU (subject to 1 year vesting), distributed among all grant recipients based on the completion and success of the proposed projects.

Sidenote: OLAS has a native dev rewards mechanism for several 100k $ per epoch (~1 month) to developers. Any agent use case minted on the OLAS registry would qualify for these rewards.

Motivation

The primary motivation behind this RFP is to bridge the theoretical and practical aspects of transaction supply chains, specifically in the areas of MEV and transaction ordering. By fostering a hands-on exploration environment, the collaboration seeks to enhance understanding, security, and efficiency within the DeFi space.

Pre-requisites

  • Proficiency in blockchain development, particularly within the Ethereum, Gnosis Chain, and Optimism ecosystems.
  • Experience or a strong interest in MEV and transaction ordering strategies.
  • Commitment to maintaining operational aspects of the deployed systems for the agreed period.

Platform

Snapshot

Voting Options

“YES”

Approve the launch of the RFP and allocate the proposed rewards for project development and completion.

“NO”

Reject the proposal.

“ABSTAIN”

Choose not to vote on this proposal.

License

CC0: This work is marked with CC0 1.0 as dedicated to the public domain.

2 Likes
2

Hello,

I fail to understand exactly what we are talking about here. Shutter encrypts the mempool such that we prevent MEV.

Are you proposing to allow AI agents (whatever that means) to access the mempool and frontrun them?

1 Like
3

Hey, thank you for your question!

No, not exactly. A key issue with the encrypted mempool is how exactly to test the economics (how much front-running can it prevent, how is end user trading P&L affected etc). Gnosis Chain doesn’t have a lot of MEV and the testnet has virtually no MEV.

OLAS autonomous agents could allow to simulate real user (and potentially searcher) behaviour and then measure the economic impact.

So essentially it’d be stress testing the encrypted mempool and simulating real user behaviour with real funds in a low stakes environment.

1 Like
4

Great proposal that could provide evidence of Shutter’s value proposition and support further development.

I would like to understand the following a bit better:

  • What are the next steps for this proposal?
  • How will rewards/budget be structured if this is a joint RfP between Shutter and OLAS? I understand that Shutter is committing 50k USDC and 500k SHU for this RfP - how much is OLAS committing to the reward pool for the RfP?
  • What would be an indicative timeline to complete this project?
1 Like
5

Thank you for the positive feedback and those questions!

  1. Next steps for this RFP are the usual steps for RFPs, so once we have sufficient feedback here, it will go to the Snapshot to get approved and is then considered live, after which anyone can submit proposals (thinking 2 weeks for dev houses to submit proposals). The next steps on the “bidder” side is up to the proposer but there are some ideas outlined in the proposal for what could be built and in which order.

  2. All of the current rewards are proposed to be committed by Shutter DAO 0x36. Sidenote: OLAS has a native dev rewards mechanism for several 100k $ per epoch (~1 month) to developers. Any agent use case minted on the OLAS registry would qualify for these rewards.

Once the more technical aspect of this experiments are set up, a follow-up proposal could be issued to Shutter DAO 0x36 (and OLAS and Gnosis DAO potentially) to fund and further incentivize the autonomous trading agents and liquidity.

  1. That’s up to the proposers who bid for this RFP and depends largely on the complexity of the OLAS stack. In my mind, a basic version similar to the OLAS BabyDegen and standardized front-running bots could be set up in 1-3 weeks.
6

Really like this, great combination of Shutter, Olas and Gnosis to build for a very interesting project. We will certainly be submitting a bid once approved.

1 Like
7

Thank you so much for the feedback and looking forward to the bid if approved!!

8

I think it is a very interesting project and like the integration of the technology along with the increased visibility it will bring to Shutter.

I am not however convinced of the benefits of the actual project with regards to long term shutter adoption. It feels like we would be funding artificial trading and mev capture in order to then test whether that artificial mev would be prevented by shutter. I am not sure how much that proves or how convincing it would be to someone who knows that there is not a large volume of mev or trading on gnosis currently. The money may better be spent funding research into the amount of MEV that would be prevented if Shutter was deployed on a chain that has higher levels of MEV currently.

2 Likes
9

Fair point! In a way, I do see this as a research project into how much MEV would be prevented on a chain with higher MEV. We’ve tried to do this via pure data analytics/research in the past, e.g. here: The Viability of Integrating an Encrypted Mempool as Proposed by Shutter Into the OP Stack, so this could essentially be an alternative approach more akin to simulation or using cadCAD. I think the proposers could take this into account and focus on the research and analytics aspect of this, i.e. essentially the dashboard and measuring part.

I’m also considering providing 2 options for scope and budget which could mitigate some of the risk of spending too much money on this experiment.

10

I am not sure this makes too much sense. I agree with @d0z3y that this artificial trading might not represent real world usage and it might fail to give any useful results or even give misleading results. I would assume that a researcher could work on this for a month or two for less money and come up with results of the same quality or even better.

I’d most likely vote against this proposal, as I fail to see what value it will add.

11

Ralph from Valory here (core contributor to Olas). This is a fantastic use case for Olas autonomous agents and we look forward to support this and other Shutter agent use cases.

1 Like
12

Hey everyone,

Thanks for all your feedback on the OLAS x Shutter DAO 0x36 proposal! We’ve made a couple of key adjustments based on what you shared and posted the proposal on Snapshot: Snapshot

  1. Reduced Funding: We’ve cut down the funds to 30,000 USDC and 300,000 SHU, keeping things lean since we’re still in the experimental phase.
  2. Bigger Focus on Research: We’ve beefed up the research aspect:
  • We’re adding a Research and Analytics Dashboard to better monitor how the project is doing and see if we can cut down on MEV on Ethereum Layer 1.
  • Proposers are also encouraged to dive deeper into how using an encrypted mempool affects MEV scenarios with a dedicated research study.

We really appreciate your insights—they’ve helped us shape the project to better fit the community’s goals! Looking forward to people voting on it and continuing the discussion!

Btw, I’ll also be speaking about this use case today on Olas Currents (on 27/06 @ 16:00 UTC) x.com

And also in two weeks at the OLAS Agents Unleashed event during ETHCC: Agents Unleashed at Brussels EthCC: Supercharging AI Agent Economies · Luma

Hope to see you there!

13

Update: The RFP passed (with solid support, 89.49% yes votes) and can now be considered active and open for applications: Snapshot

Looking forward to applications/bids!

14

I like thihs type of innovative research, but I also want to make sure how those money will be use. 30000 USDT + 300000 SHU is quit big amount for this type of activity and also why we need to pay SHU?

1 Like
15

Hey, thank you for your thoughtful feedback! Here are our main points:

  1. We appreciate your concern about the significant amount of money requested. However, we believe this is necessary to build out a robust autonomous agent system. The detailed proposals from Isotropic and Keyko demonstrate the extensive work required. Here are these proposals:

  1. It’s important to note that this vote is not about the grant amount, which was already decided in an earlier vote (Snapshot link). Prior to that vote, we had already made reductions to the grant amount based on earlier feedback.

  2. Regarding SHU: We see this proposal as having significant long-term potential, and we aim to incentivize the development team to stay engaged with Shutter DAO 0x36 over the long term. The SHU tokens would be subject to a one-year vesting period as outlined in the original RFP.

We hope this clarifies these points and look forward to your continued engagement and support!

16

Got it, thx luis for the feedback. If so I would suggest less cash more $SHU token with longer vesting for example 1 year cliff and 1 year linear release. I will vote yes but the porposal to support this direction but not for the expense.