Building with Hyperlane Interoperability Framework - ETHGlobal Bangkok 2024

[Music] hello everyone I'm Daniel I'm on the engineering team with hyperlane uh you probably know hyperlane already um if you've looked at Cross chain anything uh so we do General message passing and we do token bridging we do that permissionless which means we allow people to self- deploy and so at the hackathon today you can do and build on top of any of that uh but since intents are one of the most interesting things happening uh nowadays in the world of crypto I thought it's most relevant to look at how hyperlane can help you build intense on top of the existing deployments we have and so the agenda for today is to look at how intense compare to message passing message-based token bridging um and then see how intense can really benefit from hyperlanes architecture um and based on that we'll see what kind of bounties you're able to build so let's start by looking at message based token bridging and you know at core hyperlane is this General message passing protocol um in practice all of these message passing Protocols are used just for token bridging that's literally 99% of the volume and War routs are hyperlane stake on token bridging they're built on top of General message passing and it's a fully open architecture and framework so they work with any token standard and they work with any sort of security module uh we call that modular security and it means that if you integrate with hyperlane uh your protocol is future proof in terms of enabling you to upgrade your security as better models become available such as ZK light clients um and so let's just look at how powerful war routs are I'm going to show you a demo um owl protocol built a UI for self- deploying a workout it takes 60 seconds to do the deployment and then it just works you don't need to run any infra you don't even need a wallet to use owl protocol because they have count abstraction and it's pretty much a one-click deploy and then you can start bridging via super Bridge uh this one yeah so they have like this um website interface where you can select uh the tokens that you want to connect using the the war routes um and so you get to connect the origin chain let me start the demo you get to select the origin chain and then the token you want to bridge from that chain in this example uh they're going from the spolia testet bridging native eth to some destination chain um minting synthetic eth uh where and yeah on the right hand side you just said the token name and symbol for the synthetic but what routes are out you or allow you to connect collateral tokens as well which means you don't need to Min synthetics every single time like to wrap assets you can just uh provide liquidity and allow people to lock and unlock essentially uh and so in this example the word Prout connects three chains uh Native eth onoa and then synthetic on the other two um the user click deploy and then the transac c s are out on all three destination chains um you see the Live status there and you didn't even need a wallet interaction because their account abstraction and the really really cool thing about this is that uh after the deployment is done you get a configuration with deployment artifacts of the war route which you can then use in the super Bridge app uh which is a bridging interface and you can start bridging straight away you have a UI with the bridge you just deployed and you then select your you know origin chain the token you on to bridge um and like you're good to go you have a live token Bridge the only thing remaining to take this to production is to move ownership of these contracts to something very safe cuz it's currently owned by the hotkey in account abstraction all right um and so moving on we've seen how hyperlane has made tokening based on messages very very easy to do uh let's look at what intents are hyper Lane is building on top of that um an intent is really a user expressing uh their wish to chain um they do so by sending a transaction On the Origin chain then there's an chain actor called usually a filler that will notice that intent and then perform an action on the destination usually this means the user wants to bridge some token from An Origin chain to a destination chain and so because the filler will have balance on the destination they will send those funds to the user immediately um yeah and then the steps are essentially that number one is you sign an order that gets sent on chain on the origin chain um there's a settlement contract that takes the user's tokens into escrow after tokens are in escrow the filler can send funds to the user on the destination chain and then the last step is to reimburse or pay the filler with the tokens that were held in escrow from the user and so you know people who build intents almost behave like oh you don't need in order to uh you well do cross chain actions but the truth is that for settlement purposes you absolutely need General message passing um because the filler needs a way to prove that they sent funds to the user on the destination and so yeah General message passing will always be required and the question is how can we do that how can we make that easier uh just found it interesting that over time uh clearly intents have uh gained a lot of popularity now they're around 50% of you know cross-chain volume compared to message uh based token bridging uh and let's look at some facts about you know message based token Bridges which are warp routes and intents so War routs are slow because they need to wait on finality On the Origin change um it if you don't wait on finality a reorg could occur and so the tokens you sent initially are then you know given back to you cuz it's as if that transaction never happened so uh you kind of end up dumble spending because you receive tokens on the destination and you get them back on the origin as well uh but they're very easy to bootst strap like you've seen the demo it takes around 60 seconds to deploy you don't need inventory uh in order to fill users on the destination chain because in most cases you mean synthetics uh and it works between any chains uh because of that reason um and what's more execution is guaranteed because you can like even if there's no relayer to send your message to the destination chain you can relay your own message that will you know cause new tokens to be minted um so that's pretty convenient but then they're slow right um and that's where intents really shine because the filler in many cases doesn't even need to wait on finality uh the fee you pay to the filler uh to do the cross chain intent uh execution usually acts as this reor insurance almost so basically on average or like you know reorgs happen very very rarely so it's very safe for fillers to assume that over time the fees they get uh are greater than any funds that they may lose uh because of a reorg um so this is the big big benefit of intents and it's why they're taking off but they're much harder to bootstrap because you need fillers and those fillers need inventory um and they need inventory either in any token that you're trying to bridge cross chain or you know they're going to need to take their say funds and then swap them and only them send them to the users I'm not sure what happened here uh there you go cool let's keep going um yeah and as I said earlier we still need General message passing Bridges for settlement um but then you also need them to mint inventory like if a token doesn't already exist it cannot be held by the filler then uh it's impossible to do intense and then also as a filler if all users go from one chain to the other like um you know bridging is very imbalanced you may end up uh no longer having enough funds to fill orders so you need to rebalance your inventory and you also need likely um like a general or traditional token bridge if there's no other filler to allow you to move your funds um and so as I said intents today are generally cheaper and faster so more convenient uh than token bridg it gives users a better experience of using the app but the bad is that uh like intense protocols today they roll their own messaging protocols um and they tightly couple security so I think a very common use case is to use like some sort of optimistic Oracle um that proves that settlement happened on the destination chain and then yeah because of that you have this these intense protocol that do both intense but then they also do token bridging so they're kind of spreading themselves rather thinly and obviously they can do everything best they should focus on what you know distinguishes them from their competitors um and as a result intents are usually limited to a handful of chains um and this is where basically we start to thinking about um you know why not use a general message passing protocol to settle intense instead of uh building your own so you can focus on the intense layer um and besides that like basically just allowing us uh or other message passing protocols to handle settlement for you what hyperlane really gives you and where it really shines is that it has modular security which means that different sums can be uh settled with different security mechanisms so say there was just a single order of one of $1 that you're trying to settle that doesn't really warrant the highest level of security you can go with a cheaper uh method like maybe some multisig or some trusted Oracle that tells the protocol yes uh a dollar was really sent by a filler to this user you can now pay the filler On the Origin chain so as a protocol you can have these sort of criteria um and this in turn makes it cheaper for fillers to operate because you know I guess a less trusted security model will be cheaper to verify um so fillers will have to pay less in fees to use the intense protocol and in the end all this means user get lower feeds um yeah so kind of like to think about all the stakeholders and an intense protocol uh the protocol itself gets to pick its tradeoff that best matches the settlement speed they want to achieve and the security and the cost for different kind of sums that end up being bridged and you can focus on attracting users making it easy for solvers to integrate then for solvers um again like you get kind of abstracted from many of the complexities of message passing and you can can focus on what you do best which is manage your inventory index new intents so you can fill them quickly and then including your transactions and then in terms of users overall you get better quotes because uh settlement can happen uh well very dynamically based on the amount being bridged so it can end up being cheaper uh and it's a virtuous cycle because you start off with you know better ux uh this attracts more users which generate fees and volume and this leads to more competition which attracts solvers and because there's fiercer competition between solvers you get lower fees which attracts more users and so on uh and with that in mind we started developing an open- Source intense framework where we already have a solver that's live we have um basically a CL that you can use to spin it up contracts that you can deploy where users can send their intents um and actually I have a demo to show you exactly with that I'm not sure how good we time but um yeah you just spin up the solver and it indexes the origin chain it will look for intents as they're being broadcast and then if the solver has uh enough balance on the destination it will fill the user's order essentially um and we're trying to make this you know 7683 compliant there is no 7683 intense protocol live in production today and we're trying to you know ideally turn this into an industry standard that other intense protocols can build on top of and then add their solver and so on yeah so as you can see the intent was noticed then the order was filled on the destination chain and then the final step is settling it which means the filler gets back gets reimbursed On the Origin chain and then there's some transactions there and just because we're very short in time uh yeah let's get to the bounties so you can build anything the best projects will get the prize um out of the 10K pool if you don't have any ideas uh what I can suggest is basically two main areas one is account obstruction and the other one is intense in terms of account obstruction hyperlane supports interchain accounts where you can control accounts on remote chains from like some origin chain and you can have balance you can initiate authenticated cross chain calls uh but like smart wallets today aren't integrated with that and then yeah the other thing you can do is build around intents so the current framework um we built only supports token BR bridging but you can have call data based intents so basically you know cross chain calls uh um you can build a request for quote system uh to have better you know pricing discovery of how much it cost to send an intent in the current version of the framework users have to kind of estimate or assume how much it will cost you can also build a UI for the intense system that I just showed you and yeah really I just invite you to come chat to us at the booth for more uh cuz there's tons of things you can build and it's all open source which makes it way easier sh [Music]