WolvesDAO - $woof

Wolves DAO

WolvesDAO is… to be honest, I’m not the guy to explain it. I’m not that in touch with gaming stuff, or anything outside of the few Avax communities I frequent. But they have a nice website, and a Substack with a lot of content, which I appreciate.

A lot of people have reached out to me over the last few weeks looking for assistance launching projects, and I’ve had to turn a bunch of them away because I didn’t know them, or they seemed too sus, or frankly I just had too much work to do.

But WolvesDAO came recommended to me from a number of people I respect. So when they asked for help launching their upcoming token $woof, I agreed.

$woof

I’ve made no secret of the fact that I’m skeptical of most AI based coins. Not that it matters much to me as the deployer, the pay for deploying a non-AI coin is the same as for deploying an AI reply guy coin.

But it does seem like they have more interesting plans for this coin. Apparently they have been collecting web3 gaming related data for a while and plan to integrate it into their AI? I assume you’ll need to have X coins to access their token-gated services, or perhaps pay in the token, not clear from their docs.

Anyway, at the time I accepted this job I was stressed out from preparing for $blub; after the success of the $wink and $ket launches, pressure was on to keep ahead of advances in sniping. So I was pretty happy when they said they just wanted a vanilla launch with no shenanigans.

Circumstances change

Unfortunately, once things heat up in an ecosystem, new coins start to become a target for snipers. And we saw that happen with $albert, which got pretty wrecked by snipers even though they had anti-sniper tactics built into the launch.

I’ve said before ‘you only get one chance to do it right in web3’. A bad launch can wreck you, like dead-project-forever wreck you. When they reached out to upgrade their launch from a ‘simple’ one to an ‘anti sniper’ one, my first thought was ‘they bait and switched me’!

But to be honest, I’m glad they did. I have my own professional pride, and having a launch I oversaw destroyed would be very offensive to me.

They agreed to pay me a bit more, and although I wasn’t willing to go to $blub lengths (undisclosed at this point) I agreed to repurpose a bunch of stuff from the $wink launch. I still warned them about the risk of sniping, though.

I asked them to push back their launch date a bit so I would have more time between the $blub launch and $woof launch to prepare, test, and decompress.

Things change, again

From a technical perspective, the $blub launch went off pretty smoothly. A few minor glitches, even after all our testing, but we wrapped up our launch checklist and were done within a few hours. I moved on to preparing for the $woof launch.

While I was working on it, someone asked me some questions about the $albert launch, and while investigating I stumbled across this.

Someone testing their Twitter scraper, including CA detection inline, in images, and in a linked website. Uh oh! I warned them about the increased risk level.

Even worse, someone asked me some questions about the temporary $blub router, and on investigation, I found more worrying stuff.

People manually poking at the router. The page had already been taken down, meaning these were scripted requests. My paranoia (and blood pressure) increased to never-before seen levels. I decided to see if they were willing to pivot, again.

They were, and they even agreed to pay for the extra work required (not extortion, I swear).

Final launch planning

Given the new plan to launch on a private AMM, we opted to scrub all the ‘anti sniper’ fake LP tactics that we’ve used in previous launches. Those things just punish snipers, but don’t actually make the launch any safer for real users. Better to skip the controversy.

VQ quickly reskinned the AMM front end and made a few improvements as well.

He claimed he was going to be very lazy about it, but he’s a liar. Look at this level of detail! What a sucker.

I also added a few novel improvements on the smart contract side to attempt to dissuade snipers. Rabby ended up being a problem, but that’s a story for another time.

Other than that, the only additional hitch was that the requirement was to launch on Pharaoh CL, mimicking a traditional 50:50 LP. I had some code to do this, but out of paranoia I ended up writing a bunch of tests and extra helper scripts in case anything went wrong.

We wrapped up launch planning, and I did a bunch of dry runs with 12 hours to spare. I registered all the domains and set up blank pages to make sure there wouldn’t be any DNS issues this time.

Launch

I absolutely despise people who give multi hour launch windows and then launch near the end.

I was about 20s late to start, but I wrapped up pushing everything and doing a quick test buy/sell on the router at about 10:02. Thanks to excessive testing, things went very smoothly.

Then I hit my first hitch.

The airdrop was to over 3K users, so out of extreme paranoia I had written a custom disperse contract that would refuse to airdrop to the same user twice. This came in very handy when the airdrop didn’t complete successfully.

The issue turned out to be that I had just slightly underestimated the amount of gas required to do all the airdrops. I put 5A in the deployer, we needed about 5.5A. Whoops.

After filling it up with another 5A the script finished the last two drops just in time for me to migrate the LP to Pharaoh at 10:12. I was finished cleaning up the deploy (transferring LP tokens, ownership, dex registrations etc) by 10:30.

One last hitch happened when the Pharaoh UI lagged on finding the new LP as a place to trade against. The aggregators (e.g. Odos, 0x) actually found the LP about 20s faster than the native routing somehow.

Final thoughts

The launch went pretty smoothly, compared to other launches. But I’m not convinced that this will be the case in the future, fighting snipers is a cat and mouse game. The stuff I put in at the last minute only (visibly) blocked one sniper for about 30A of buys.

However, a guy I know from a few years ago who is big on MEV and sniping on Avax (and other chains) did reach out to me to tell me some of my efforts recently had been at least partially successful. Based on some stuff he told me, an unknowable amount of bot activity was prevented; since you can simulate TX before sending them (and abort failures), you can’t use what’s reflected on chain as the final word on what people tried.

As clients go I can recommend WolvesDAO. They were very professional and organized, paid me prior to me even asking for payment, didn’t complain when I asked for pay for VQ to do the unplanned frontend work, and proactively offered me a bonus for the work. Was great working with them.