The Best Stack For Building Ethereum Apps And How To One Shot Bangers I Austin Griffith

[Music] All right. What's up? I'm Austin Griffith. Uh, like Pascal said, I work for the uh I work for Ethereum at the EF. My goal here is teaching folks how to build on Ethereum, and we're going to speedrun through building with the best tool out there, Scaffold.

Uh if you want to get started, there's lots of material ethereum.org. It's a great resource. If you if you have a uh a university, go join us at uh ethereum on tour. It's tour.

ethereum.org. Uh if you are a developer getting started and you know how to build some things, but you're like, "What do I build?" You need to go build a whole bunch of things. Go to speedrunthereum.

com and get started building. If you uh want to know, want to just build some things, that's what you got to do. That's what we're going to do here with Scaffold ETH. So, here we go. We're going to get started with Scaffold ETH uh right away.

You can ask questions in the chat. Let me see if I can even find the chat. Oh, yeah. I got it. I got it here.

Here we go. Feel free to drop questions in the chat. I'm going to drop scaffold.io in the chat. This is what we're going to use.

We're going to use a tool. We're going to try to build something uh in 30 minutes here and uh hopefully deploy something to production. So, uh let's get started. So, Scaffold ETH uses MPX create ETH latest. We're going to paste that in and we're going to create uh I don't know.

Let's let's see what is this. ETH online 2025. Yeah, there we go. Let's let's create this. So, uh here's your first kind of choose your own adventure, right?

You have either hard hat or foundry. We we use either one of these. Uh whichever one you like the best. Um if you're using either one of these, it's going to be all the same commands within Scaffold ETH. So you'll do yarn deploy and it will deploy whether you're using hard hat or you're using foundry.

Um I think a lot of the cracked engineers like to use foundry because you're writing your tests in solidity and it's a little bit faster. Uh I think there's a lot of older projects that are using hard hat and also um it's I don't know maybe a little friendlier for deploying. I don't know some some people like to use hard hat still. We have both. You'll find a lot of things like this in the ecosystem where there's multiple options and that creates uh diversity and uh it works great for a decentralized ecosystem.

My WETH node over here died the other day and if we were only using just you know Ethereum would have been in a lot of trouble but thankfully we have GE and other uh consensus and execution layer clients and a strong diversity leads to a stronger ecosystem. Okay, so scaffold ETH starts out with this one command and uh you CD into your build and you're going to need to fire up a few things. You'll run a chain which uh is Anvil or the other one. Basically, you get a local host uh blockchain and it has a bunch of addresses that allow you to uh you know send Ether around, do whatever you want. It's just like a little copy of Ethereum.

And then you can yarn deploy and that's going to deploy a uh placeholder smart contract. And we'll also need to start up a new window for yarn start which is our front end. And there we go. That's the whole setup. That's how you get started.

You have your chain, you have your front end, and then you can kind of deploy over here uh in this terminal. And we'll come back here and we'll deploy a bunch of things uh in this terminal. But feel free to ask questions in the chat if you're if you're joining along. We're going to let scaffold ETH load up here. And we're going to use cursor today.

We're we're using a lot of AI, right? So, I'm going to open up cursor. And that's going to be over here. Let's see. How do I shrink this?

There we go. I'm going to shrink this and put it over here. I don't know if it's even useful. And scaffold comes up. This is what scaffold looks like.

This is your front end. You've got um burner wallets and a faucet. And you got you have this debug contracts tab. And this is where you'll probably spend a lot of time at first. Uh hitting that.

Let me see if I can get some light here. Let's go. Let's get some Which button is it I'm hitting. Oh, there we go. Really bright.

Okay. So, uh let's see. We've got our debug contracts page. This is where we can poke at our smart contract and scaffold the ETH foundry contracts. Let's get in here and look at our starter contract.

So, scaffold comes with a starter contract. Uh in this case, it's kind of a greeter contract. You can see here there's a set greeting function and it updates some string storage, right? And so if we go over here and look at it, here is kind of our interface where we can interact with that contract. And we can see we could set the greeting to hello world.

So this is going to fail. And when I hit this, it's going to say, hey, you don't have any gas. Oh yeah, I've got to go to the faucet and grab some gas. Right? You'll also notice that as I make this transaction, there's no pop-ups for chain ID or moving me over to the right network or anything like that.

Uh, we don't have to sign any transactions because we're going to use burner wallets. When I hit send here, it just makes the transaction. You'll notice there's a nice little counter. Watch. I can make that counter click, click, click, click, click, and go.

So, this is really nice for testing when you're on local host. You can just hit buttons and interface with your smart contract. Poke at it, get it to do the things you want it to do. I always like to, let's see, we've got this total counter here. I always like to make a new function called like count by three or something like that.

And no, not return three. What we want it to do is do total counter plus equals three. something like that where every time you run this function, it's going to increment your total count by three. Uh, it's not liking that. Let's let's ship it.

See what it says. If it doesn't deploy, we'll we'll let uh AI fix it. So, let's redeploy our contract. And we're going to add a reset there to make sure we force a new contract out there. Oh, yeah.

It's not letting me do that. Uh oh, we don't want to return anything. Needs to not be pure. There we go. It just added some extra things in there.

That's why. Okay, here we go. Let's redeploy our contract. So, this lets you kind of tinker around. It lets you quickly tinker with Solidity, deploy a smart contract, and then interface with that smart contract.

We we can see that old address, now we have a new address. It kind of hot reloads, and now we have this new count by three button, right? And when I click that, there we go. Our counter increments by threes for some reason, right? Now you can imagine we you know make this payable and we maybe come up with some uh UN 256 public price and set it to you know 0.

001 001 ether. And then maybe we require that they pay the price. And then maybe we do something really cool where we say price equals price uh times 104 divided by 100. Right? You kind of have to do this numerator denominator thing because you can't use decimals.

Uh EVM doesn't do floatingoint math. You learn all these things as you start tinkering with solidity. But now we have some count by three function that you have to pay to use and there's some price and that price will go up on a price curve. Just kind of randomly tinkering with solidity. We're playing around here.

Now we have to pay money, right? We have to pay this much which also is 0.001 ether. And you have to take that times 10 to the 18. And you'll notice that as we hit this uh our price goes up.

If I try to pay that again, it's going to fail. So, we have to pay, you know, it it's going up by 4% each time on a nice little price curve. And we can see that price going up. And also something interesting is happening here. Our contract is collecting money, right?

There's money loading up in this contract. You're literally programming money. So, this is uh how you get started with Scaffold ETH. Go to scaffold.io, run that command.

You'll have to bring up your local blockchain, your front end, and then your yarn deploying all all you want here with with any changes. You've got burner wallets and smart contract uh kind of hot reload adapting to your contract. As you add new things, this debug page changes. Then eventually you move over to this kind of frontend area here and you kind of edit out your front end to talk to your smart contracts. Maybe you're going to talk to an existing smart contract.

You add them into like your external contracts in in scaffold ETH and then you could maybe like I don't know have something that like puts money into unis swap or something with a button push here. All of that is good for tinkering and getting started. Uh what I want to do today I think the title of the talk is the best tool I should show the docs real quick. Um the best tool for building apps on Ethereum uh is scaffold ETH. It's great for building apps quickly.

There's a video here that watches me. I think I build a whole app in eight minutes. Very similar to what I'm doing here. Um, if you want to interact with your, we were talking about external contracts, you could go to the docs, right, and and find this. Let's let's paste the docs into the chat just so folks have it.

Boom, there it is. Uh, if you're interacting with your smart contracts, though, you're going to use these two important hooks. This is the hook that you would use to read from your contract. This is the hook you would use to write to your contract. And there's nice little examples here you can copy paste into your code and get started.

So that's how you would do it the manual way. Today we're talking about uh the best tool for oneshotting apps. So we're going to use uh Claude and Cursor and we're going to try to uh oneshot some kind of an app. Uh I've been uh a lot of times on the tour. I think I already shilled this, but let me shill it one more time just because we're shilling.

On the tour, I do a lot of uh where uh talks where we build a voting app and we show how you can build an app where people can vote and I show a centralized example and then I rug the simpleized example because it's uh it's centralized. I can rug it and I can change the vote and I can censor it. And so then we build a a smart contract version that you can't rug which is kind of fun. Uh but uh let's build something interesting here. I was thinking about FOMO 3D.

If anybody uh remembers FOMO 3D, it was this app that you would uh you had to like keep pushing the button and if you the last person to push the button when the time ran out get paid got paid all the money in the contract. And that was such a funny uh mechanism. It was basically a bounty to break the the chain. I feel like we should try to oneshot something like that. So uh here's what we have.

We have uh scaffold ETH up. It has uh our smart contract here, but there's also a cursor rules file. So, it knows how to make the front end. It knows how to make the back end. Uh we're not gonna let's let's not even mess with our contract at all.

We're going to start with like a complete blank thing. No context, not talking to anything, just the cursor rules. I'm just going to try to oneshot into scaffold eth and see if it builds it. I don't know if it will. Let's see what happens.

Um, but let's build an app like FOMO 3D where there is a single big button and a countdown. So after it deploys, let's see, the counter starts counting down. for let's see from like five minutes or something like that, right? Well, it's not it's going to be hard. We're we're going to hardly have any time to test this if we put it out there, but uh let's say 2 minutes.

Uh if it gets to the end, anyone can run a function that pays all the money in the contract to the last person who hit the button, right? So, uh, so when you click the button, this is a terrible description. I hope I can figure out. Click the button. Uh, the time, uh, adds two minutes and keeps track of the last person to click the button.

Right? That's that's the whole smart contract. The front end should just be a button and a countdown. Something like that. Let's see what happens.

We're using Claude and cursor here. Uh let's see if it's able to write our smart contract. Yes, FOMO 3D. So, this is FOMO 3D in 2025. Uh you know, we don't have millions of hours spent pouring over the Solidity, making sure it's perfect.

We probably should and AI may not be at the point where you should be writing smart contracts with it yet, but I think uh we're going to try it today. Maybe it will have a giant bug in it. Uh if we have time, I'll put it out on a public network just to tinker with it, but let's see where we go. Okay, so it looks like it's already writing a contract for us. Let's go look at it and see how it looks.

There is Yeah, so there's the last clicker. There's the time increase. There's the how much it costs. That's kind of expensive. Maybe we should make it a little cheaper.

Uh let's see. So, uh when we deploy it, it will have it will start at 2 minutes. When you click the button, it's going to probably add the time increase to to the game end time. We're keeping track of total clicks for some reason. That's going to cost money.

We probably should put this on an L2, not uh Ethereum mainet. Uh and then you get to claim your prize. There's time remaining. Get pot size. Let's look at claim prize.

Let's look at the require statements, right? We're making sure the game hasn't ended yet. We're making sure it's after the time. The last clicker isn't the zero address. Then it sets it to true.

Gets the prize money. How does it send it? I get out of here. And does this, right? I think you could grief this by having uh this be a smart contract with uh a a receive function that does some extra execution.

So there's a lot of ways to grief this. But here we go. It looks like it made FOMO 3D for us. And when we click, hopefully we see this time uh go forward. I don't see the time counting.

Let's see what's going on here. Okay, I feel like the timer is not working quite right. Maybe it's still working on something. Cuz as we click this, the time should click up by Oh, wait. We need to deploy our contract.

Yeah, yeah, yeah. Let's Let's deploy. Oh, not reset one. Uh, compiling. Compiling.

No changes found. Let's yarn deploy reset. Let's make sure we get our contract deployed. All right. And then we could go look at it in the debug contract page.

It's still loading our front end up here. We got to wait for this to come in and we'll see if we made FOMO 3D. Yeah, there we go. So, that's the whole thing. There's just a clickable thing.

You pay some cost. Let's actually lower that cost, right? Let's put it to 0.001. Uh, which is still what is that like 40 cents?

Something like that. Let's Let's go with that. Let's go with that. Let's yarn deploy the new contract. Make sure we get a new contract here.

Make sure the price goes up a little bit. Or goes down a little bit. There we go. Okay. Here we go.

Let's look at it. So, here is everything. If I click this, I don't see the price going or the time going up. So, you may So, we didn't oneshot it. We failed.

We failed the oneshot uh thing here. Uh let's see the timer. Wait, wait, wait. Let's go check and see if there's a time remaining. Let's let's add that in, right?

Uh should be a time remaining function that tells Well, yeah, it knows what that is. and the front end. If I click the button three times. Let's see if that works. Because what should happen when you click the button is it should increase the time.

Ah, see. Yeah, there's not. It should actually be game ends plus it should be this right here. Oh, no, not that. It should be game end time plus time increase.

Not. Yeah. Yeah. Hold on. Yeah, this is a fail on the oneshot.

I think maybe I didn't explain it very good very well, but let's see. We want it to be something like this where it's not the end. It actually like increases the time way out. If someone clicks it three times, it should increase the time way out. But maybe not.

Maybe that's not even how FOMO 3D works either. Let's just see. Let's see what it does here. Terrible oneshot. Okay, we got we have 10 minutes left though.

We need to put this out into production so we can play with it. Uh just showing that it doesn't take long here. You're you're kind of prompting your uh you know nudging it along. Sometimes if it doesn't one-shot it perfectly. Uh let's make sure it's still doing some frontend stuff.

I don't want it to deploy. I want to do the deployment myself over here. Let's see. Let's make sure we get a new contract here. Uh oh.

Uh oh. What was that? Oh. Oh, this guy was doing a deployment at the same time I was doing a deployment. So, we got some nons errors there.

There we go. Okay. Fresh contract. There we go. Looks good.

Uh, time is counting down. Oh, weird. Look at that. If I hit reload, it always restarts to There's There's definitely a front-end bug. Really?

Oh, wait, wait, wait. I mean, that looks good. Three minutes and 40. Okay, no, it's actually working.

If I wait a little bit and hit reload, I'm getting a fresh number. If I click this, it's going up. Okay, so now we need to test real quick if like we can get the money out of it when uh the time Let's just set it at 0.1. Uh, is there a button to get the money out when the time runs out?

So, as you're building things, you're programming money, you need to think a lot about incentives. We could set this up. So, I mean, the person who gets the money is going to be incentivized to push the button, but whatever you build, whenever you need an automated mechanism, you need to be thinking about incentives and build those incentives into your smart contract. So someone out there, whoever can click the button to roll your logs or uh you know increment your system, move it into a different state, but they have to be uh incentivized to do so. And you have to be thinking about incentives as you're programming these mechanisms.

Okay, here we go. Uh let's see. I turned it down to one minute. And let's deploy a fresh one. And let's actually we should do like 30 seconds.

Let's let's do like 20 seconds. We don't have time here. Let's deploy another one where it will start with 20 seconds. I'll click the button. It'll put it out to 40 seconds.

Uh the money will be locked in there and then hopefully it will run out of time and we'll be able to sweep the money out and prove that it works. All Oh yeah, there is a nice claim prize button there. Very nice. It kind of flashes. That's pretty ugly.

If we were building this forever, we would want to make that look nicer. Okay. There we go. So now we just have to sit here and wait for 30 seconds. Let's get started on the deployment.

If we're going to put this out on a live network, we probably need to Let's just do deploy- Oh, actually, wait, wait, wait, wait. Do we still have those console logs in there? We don't want console logs. Uh yeah, yeah, yeah, yeah. We want to pull this out.

We don't we don't want any console logs. Uh this stuff, right? We don't want to put console logs out on mainet. That's just a bunch of trash put on mainet. Let's make sure this still Oh, no, no, no.

Let's I don't want to do that yet. There we go. Let's claim the prize. Make sure this works. So, there's 0.

01 in the contract. If we go look at the debug contracts page, we see that there's like 41 cents in there. Uh, if I hit claim prize, oh no, it says game still active. That's not good. Uh, why is that?

Let's see. The claim prize game still active. So the block timestamp has to be greater than the game end time. Oh no, did I redeploy? What happened there?

This is the worst worst oneshot in history that usually you can just like prompt the thing and it will build like I do always do this with a voting app. I've done even more complex apps than this. I don't know why. Let's make sure. Am I using the right model?

I'm usually using cloud 4.5 sonnet. is pretty good. Uh something happened here though where it thinks that the game was still going. It's worth the extra 40 seconds to make sure we don't put just junk out on mainet.

Uh okay. So it's that 30 seconds. If we look at the game timestamp. So whatever this time stamp is needs to be less. Yeah.

Hopefully we don't get game still active here. Let's see what happens. We might though. There might be a bug in this code. Uh, I guess we'll find out in just a second.

Wait for this to countdown. Wait, why is it 30 seconds? Uh, I feel like the timer is not working. Oh man. Let's see.

Uh, and if I try to claim the prize, it's going to say game still active. I may need the console log so I can figure out what's going on here. What's wrong? Uh, I feel like Let's go look at the code real quick. Let's Maybe we have to do something manually here.

block timestamp game end. It starts at time stamp plus increase. That's I mean that's pretty good. That's what we want. I feel like the front end is just failing here.

This guy here. There's something wrong here. If I hit reload, it's going to go back to 30 seconds. Yeah. Uh yeah.

Oh. Oh, you know what? I know what this is. Okay, we're on local host and there's no transactions happening on local host. So if I make a transaction here, it's going to notice that the time has passed.

That's the whole problem. Okay. Now, if I reload, it's at zero. Okay. So there's no transactions triggering on our local network.

And so the time stamp is always whatever the last transaction we made. So this works. Okay. We're good. We're ready to go.

Let's go to production. We only have five minutes left. So what we want to do is deploy to uh some network. I would say which whoever let's go look at whoever is at the top of L2 beats right now. Uh, it was Arbitum last time I checked.

They're there. I usually go to Arbitum or base, but let's deploy deploy to arbitrum. Uh, let's see. Arbitum, I think. Let's just use this dude here.

So, normally it's going to create a key store, a key pair, and you have to put some money into it. I think I just have created one recently. Let's see if it has any money left in it. If not, we may need to load up some money. Yep.

Okay, it doesn't have any money. Okay, let's yarn generate one then real quick. We're going to create a generator account. Uh we're going to generate our own account. We're going to call this aa uh we're going to put in a super uh secure password.

Now, if I do yarn account, I will get that account back and we'll say number two there and I'll put in my super secure password. Okay. And now we have some money. Now we need to send some arbitum ETH into this dude, right? Let's open up my wallet here and let's hit send and let's send to this dude's address.

And we want to send some arbitum ETH, but not very much. That should be enough. Hopefully 41 cents is enough to deploy this smart contract. All right, there we go. Now let's yarn deploy to arbitrub and let's see if our uh let's see number two there with our super secure password.

Please work second try. Here we go. We're deploying our contract live to arbitum. Wow. It can't be more than 41 cents, can it?

How much does it cost to deploy on arbitrum? Uh, maybe I didn't send. I did. Uh, $4. No way.

No way, dude. I have no idea. Maybe there's a gas bike right now and it costs a lot more than I think to deploy to arbitro. Uh, let's do a yarn account and pick our account. And so, it's number two there with that password.

Now, does it have some arbitrum? Yeah, it's got plenty. Okay, let's try one more time. Yarn deploy network arbitrum. Select number two there.

Enter our terrible password. Please work fourth try. There we go. There we go. Okay, our contract is deployed.

Now, we'll need to go to the scaffold config and we'll need to tell it that we deployed to So, we're not on Foundry anymore. We're on arbitrum. And there we go. Now our whole app is going to reload now and it'll be on arbitum. Actually, it's going to be so quick I won't even be able to get the Yeah, it's already out.

Okay. So, we need to redeploy again, but we need to change our time back to two minutes, right? Or maybe we don't have time. Okay, I'm going to I'm going to set it to five minutes. So, uh, you pay 41 cents and it goes 5 minutes and then if you let it end, you'll get the money back.

Let's let's deploy one more time. Hopefully this works. We got number two there. Super secure password. Okay.

And then hopefully we get our contract. There we go. This should reload. Awesome. Okay.

Okay. Then if I do yarn IPFS, we're going to put this front end out on IPFS and I'll be able to copy and paste the address into the chat and then y'all can go to this funny website and click this button here. Connect your wallet. I'm going to put my first 41 cents in here and Oh, wrong network. Bumps us over.

Click here. All right, I'm going to put 41 cents in and it should take the time out to Oh no, our front end died because I'm deploying the IPFS stuff. All right. Oh no, build error in IPFS. Let's see.

Is this up? Let's bring this back. I may have to push it to Verscell. It's way cooler when you can deploy the app to IPFS. But maybe because I'm trying to do two things at once here, it's not liking.

page block explorer. Yeah. Okay. Let's go. Let's go hit the thing.

I think this is the end of the demo. The oneshot was barely successful. We were able to create a FOMO 3D like app on Arbitum Live uh with not a single prompt and one step. It looks like that transaction did go through. Uh you can click this and it's going to push things forward five minutes.

Uh and once this time runs out, uh whoever was the last person to pick it, which is the current winner here, can come withdraw their funds. Let's try one more time to see if I can get this to push to IPFS. If not, I'll have to yarn versel and it will take a much slower time getting uh to versel and I can paste the uh the app out to my Twitter or something like that. Do you need API? You do not need an API key to deploy to IPFS.

It's actually using uh BG IPFS which is our own uh panning service uh right here. Yeah. So, BGIPFS is what's used under the hood for scaffold ETH. And uh let's see. Did our Oh, wait.

Oh, wait. I'm going to have it. It's going. I think it was because I was doing too many things at once there. I was running my front end and I There we go.

There we go. Our page is live. All right. I'm pasting it into the chat. Someone Someone else go push the button.

You have to pay 41 cents and you're definitely going to lose your money. This is a terrible, terrible, terrible app. But, uh, this is just an example of how you can quickly build a money Lego on Ethereum or an Ethereum L2. Uh, here we go. I'm going to click the button one more time.

Pay another 41 cents. Uh, here we go. And it will, you definitely don't want to click the button when you're already the current winner. And it's just going to push the time back more. So, now it's 17 minutes.

sometime 17 minutes from now if no one else has clicked this. I'll be able to get in here and get my money out. But also, you can uh you can go click the button too if you want at this address. Let me let me put it into ETH. Oh, whoa.

Get out of here. Let me put it into ETH build and I can have a QR code so then anybody can scan this. Let's go to ETH.Uld. Let's go to build.

I'm going to get a QR code for the website so you can get this long uh link out. There we go. Okay, there's a QR code. There's also uh that's that there's no money in that. That's the thing where you can just click it and it generates them.

Uh but check out that QR code. If you want to go play my crappy FOMO 3D clone that I built with almost a oneshot with Claude and Cursor and Scaffold Eath, you can just scan this QR code here and uh wait for the time to get down toward the end and then click the button and then if it waits five minutes and no one else clicks it, you can claim the money and win 40 cents or something there. Cool. There's the app. It's at 15 minutes now.

Thank you. Uh thank you for playing. Go use scaffoldeth at scaffoldeth.io and uh build something cool. Thank you.

[Music]