Objective: Implement a framework for executing Lido DAO’s on-chain governance voted-in actions forwarded from Ethereum to outer networks to minimize or exclude involvement of multisigs, committees, and intermediates.
Context: For the recognized wstETH deployments on Layer 2 (L2) networks, Lido DAO currently utilizes established processes that employ canonical bridges, ensuring security congruent with the underlying rollups.
Proposal: This discussion aims to explore the adaptation of the governance infrastructure (a.DI) developed by BGD Labs to forward Lido DAO governance decisions (originated from the Lido DAO Agent contract on Ethereum) to external non-L2 networks, such as BNB Chain, to mitigate risks of third-party bridge services misoperation by utilizing bridge aggregation for messages.
Abstract
The a.DI system by BGD Labs serves as a cross-chain communication layer designed to facilitate secure interactions across different blockchain networks with minimal exposure to bridge-related vulnerabilities. This system is instrumental in AAVE’s governance v3 as a secure message bus and has undergone multiple security audits and formal verification rounds.
However, a.DI doesn’t contain a built-in execution module out of the box and requires one to be implemented supporting the interface needed. It’s suggested to derive BridgeExecutorBase for the governance motion payload format to be compatible with the one used for L2 networks forwarding by the Lido DAO.
Scope of pilot implementation
Immediate goals
Extend Lido DAO governance functionalities to pass executable decisions on BNB Chain
The proposed contractual solution shaped as a LIP, tailored for Lido deployments, is under audit by MixBytes(). Lido contributors anticipate sharing the comprehensive audit report on this forum shortly.
Testnet
Testnet deployments connecting the DAO Agent on Ethereum Sepolia and CrossChainExecutor on BNB Sepolia Testnet (Anchored to Sepolia) are available here. The deployments resemble the proposed adapter configurations.
Further details will be included in the corresponding snapshot proposal(s).
Feedback Request
Community feedback on the feasibility, security concerns, and any potential improvements to this framework to ensure robust and decentralized governance decision forwarding from Ethereum to diverse blockchain environments would be highly appreciated.
Ernesto from BGD Labs here. Really exciting to see the proposal to adopt a.DI into Lido!
As an update of the other active instance of of a.DI for Aave, it has been running in production for Aave since December without any problem, enabling all the stages of a quite advanced multi-chain governance, including the specific needs of Lido: forward governance decision to any destination network.
Additionally, we have been doing continuous improvements to the infrastructure, like:
We generally agree with the the need of this kind of module to facilitate the governance execution to other chains and utilizing a.DI that has been running in production for Aave for a while makes sense as it reduces dependency on a single cross-chain protocol.
However, we’d love to understand more on why those 4 protocols have been chosen as a component of the Lido version of a.DI; CCIP, HyperLane, LayerZero, and Wormhole.
We suppose it is because of the original a.DI’s design, but would you consider adding more bridges in the future, or maybe replacing any of them? It’s very promising to see more flexible setups with additional considerations to which components Lido will apply once the continuous improvements by BGD Labs progresses.
Really appreciate your input, and thank you for pointing out the monitoring dashboard; it looks pretty neat and super informative.
By the way, MixBytes() noted the uncompromised quality of the a.DI codebase during their audit. We will share the final report here and suggest attaching it to the cumulative portfolio of security checks.
In an ideal situation, the other adapter by Axelar might also be plugged in to get the quorum of 3/5; however, it wasn’t yet available at the time of the full-scope audit. Nevertheless, the setup allows for adding or replacing bridge adapters. That’s why I’d suggest considering it in the future based on the cross-chain governance demand, token standards adoption and upgrades, and bridge providers’ evolution.
Worth noting that the governance forwarding is not for regular use but mostly for future-proofing the cross-chain token and protocol parts.
The a.DI setup for forwarding Lido DAO motions to BSC has been successfully deployed to the Mainnet. It connects the DAO Agent on Ethereum with the CrossChainExecutor on BSC. The deployment uses the proposed adapter and quorum configurations.
