Gray retweeted
Abracadabra, magic internet money, a bit too “experimental” DeFi project dying out. Finally.
The chain is a harsh mistress. Only cybersecurity matters.
”On the other hand, the total supply of MIM not held by protocol addresses amounts to nearly $22m. In other words, the protocol sits on approximately $21m of bad debt, and MIM’s effective backing is below $0.04 (>95% unbacked)”
Short history:
- Abracadabra launched in 2021 as a DeFi lending protocol.
- Users deposited crypto into vaults called Cauldrons and borrowed a dollar-pegged stablecoin named MIM.
- SPELL was the governance token.
- Daniele Sestagalli was the public founder and was also behind Popsicle Finance and Wonderland.
- A co-founder used the name Squirrel.
- The project marketed itself with wizard memes and a community called Frog Nation.
- At the peak of 2021 MIM was one of the larger DeFi stablecoins, with a market cap in the billions.
- SPELL became important in the Curve wars because Abracadabra paid large incentives to keep MIM liquid against other stables.
- The first major scandal hit in January 2022.
- Investigator zachxbt showed that Wonderland treasury manager 0xSifu was Michael Patryn, co-founder of the collapsed exchange QuadrigaCX and a man with a prior fraud conviction.
- Sestagalli admitted he already knew and had kept working with him.
- TIME, SPELL, and ICE sold off together.
- MIM slipped off its dollar peg as liquidity left Curve.
- Sestagalli later said the Wonderland experiment was over.
- Many people never trusted the project again after that episode.
- Abracadabra also accepted Terra’s UST as collateral.
- Users deposited UST, minted MIM, and looped the position.
- When UST collapsed in May 2022 that collateral was nearly wiped out.
- MIM depegged again in June 2022.
- The team said Terra left about twelve million dollars of bad debt.
- The peg later recovered for a time, but MIM had shown it inherited the failure of whatever sat under it.
- Security failures then repeated.
- In January 2024 a rounding bug in Cauldron logic was exploited and about six and a half million dollars was drained. MIM traded in the mid-seventies before recovering.
- In March 2025 a flash-loan attack on GMX-linked vaults on Arbitrum created phantom collateral and extracted about thirteen million dollars of MIM.
- In October 2025 deprecated Cauldron V4 contracts were exploited. A batch function let the attacker borrow and then reset a solvency check. About 1.79 million unbacked MIM left six vaults, was swapped to ETH, and was sent through Tornado Cash.
- Combined losses from those three incidents exceeded twenty-one million dollars.
- The team often said user deposits were not directly seized and that the DAO would buy back or burn MIM.
- Holders still saw a stablecoin that kept creating unbacked supply when old code broke.
- In June 2026 MIM fell through the seventies and eighties and then to about forty-nine cents.
- The team raised interest rates on every Cauldron, including dead markets, to force repayments and shrink supply.
- Curve bribes and liquidity rewards were paused.
- A small pool refill earlier that month did not hold.
- Circulating MIM was still around a hundred million dollars on a protocol that could no longer defend a dollar.
- Stewardship of the project changed that same June.
snapshot.box/#/s:abracadabra…
Gray retweeted
10 Legendary Crypto Twitter Interactions
1. You're telling the CEO of IOHK, founder of Cardano and Ethereum to use the support email?
Gray retweeted
Live evidence for the @aave x @KelpDAO x @LayerZero_Core governance war room on second-order risk.
@avantprotocol is currently running a single $156M sUSDe looping position on Aave Mantle.
$140M USDT0 + $20M USDC borrowed against it.
Net APY -135% (net of sUSDe embedded yield the pain is less, but still deeply negative). Health factor 1.02. Borrow power 100%.
This isn’t a rounding error. It’s one of the largest single positions on @Mantle_Official, sitting at liquidation threshold, eating deep negative carry, structurally dependent on Aave Mantle staying unfrozen indefinitely.
Under the L1-backstop / L2-zeroed proposal, forced unwinds of positions like this are the transmission vector back to Core via sUSDe depeg pressure on thin L2 secondary liquidity.
You can believe this position is fine in isolation.
You cannot believe this position is fine in isolation AND that L2 zeroing has no second-order cost to stakeholders like @avantprotocol. Pick one.
More quants, fewer lawyers in the war room, for the sake of this industry. 🫡
Replying to @aave
@aave @StaniKulechov
One modeling angle worth stress-testing before the L1-full-backstop / L2-zeroed proposal goes to vote: Second-order effects.
@Mantle_Official carries material @ethena USDT exposure. A haircut to Ethena from L2 zeroing could push USDe collateral on Core to reprice, looping the exact contagion the plan was meant to contain.
@gdog97_ what’s your read on this path?
L2 zeroing isn’t a free option.
It has quantifiable cost paths back to the market you’re trying to protect.
Setting aside the moral one, this is also a modeling problem, not a legal one.
More quants, fewer lawyers.
Gray retweeted
Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software.
It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans.
anthropic.com/glasswing
when I first joined Yearn in 2021, people were begging us to build UST strategies. In 2022 we were earning like 2% onchain, not too far away from where we are right now, and people could "get 20% on UST" so we lost over $2B TVL overnight
i've run into more than a few people who i recognize from back then, who were once exposing their customer to UST, and are now working on the institutional side of crypto
Replying to @VonNeumMeme
both those were ages and ages ago.
yearn immune bc they’re basically the only motherfuckers who legit don’t trust anyone and peer review everything. they act like everyone’s DPRK to the point where not even DPRK is DPRK lmao.
Gray retweeted
A clear exploration of how stablecoins are evolving into a global financial utility and the infrastructure required to keep them secure, liquid, and accessible.
Moderated by @Tanzeel_Akhtar
Speakers:
🔸Sam Elfarra | Community Spokesperson | Tron DAO
🔸Daniel Lee | Head of Web3 | Banking Circle
🔸Marcelo Sacomori | CEO | Braza Bank S.A. - Brazil
Gray retweeted
🚨 🚨 🚨 489 #BTC (51,049,887 USD) transferred from #Kraken to unknown wallet
whale-alert.io/tx/bitcoin/98…
Gray retweeted
The wait is over—OpenEden is officially opening the doors to our @discord!🟣
Join our growing community. Explore our ecosystem. Stay in the loop on all things #RWAs.
👉discord.gg/FuPtnSJmke
Gray retweeted
liquidity, users, volume, mindshare
all flowing to $S
Gray retweeted
EXCLUSIVE: Curve Finance founder raises funds at $50 million token valuation for new project Yield Basis theblock.co/post/341554/curv…
Gray retweeted
1/ Excited to launch EtherFi Season 5 point tokens!
With liquid @ether_fi pts, users can earn real yield by turning their points into profits without any lockups.
Let's see what you can do 👇
- point selling, lowest risk. our highest supported etherfi pt stream (3.5x boost) is providing a 0.22% APY lift, e.g. on @0xfluid weETHs/wstETH Looping, which is earning net 55% APY on ETH 🔥
- point buying. kpEF-5 point tokens are currently valued at $418m implied FDV, assuming 8.9T terminal point supply (as in szn 4). $ETHFI is trading at $1.2B at the moment
- providing $ETHFI / EtherFi point token liquidity to Rumpel's Univ3 pool to earn trading fees + Rumpel Straw 🌾
Want access?
Follow + RT + reply "etherfi" to this tweet for an access code (via DMs)
Gray retweeted
Spoke with a bunch of "DeFai chain" teams, turns out many don’t really understand AI or DeFi lol.
🤔 If the goal is verifiability, why even make a chain? Chains offer immutability, with verifiability as a byproduct, but agents can achieve verifiability on public cloud infra too.
🤔 If the goal is liquidity, an AI agent won’t give you an edge. It’s about designing productive use cases that cater to capital allocators’ risk and yield expectations, not AI gimmicks.
🤔 If the goal is security, AI agents will likely make DeFi less secure—fuzzy inputs lead to fuzzy outputs, introducing unpredictable attack surfaces.
Agentic products will expand upper-funnel distribution while making long-tail SaaS obsolete. DeFi, at its core, is an autonomous "SaaS", but its current distribution is broken.
I don’t doubt the organic and intrinsic alignment potential between AI and DeFi, but most teams need to refine fundamentals and answer the WHY before chasing the hype
Gray retweeted
1st time BTC hits $100k, ETH is at $4k.
2nd time BTC hits $100k, ETH is at $3.9k.
3rd time BTC hits $100k, ETH is at $3.7k.
4th time BTC hits $100k, ETH is at $3.4k.
5th time BTC hits $100k , ETH is at $3k
Conspiracy theory, but what if the underworld created this movement where the more Cross(es) there are in circulation, the more our man J.C is suppressed...
Imagine a vampire injured by a silver bullet and people carry silver bullets everywhere. Surely it's a form a deterrence.
Gray retweeted
The @Soneium situation is a good live demonstration of how launching an ethereum L2 is great for businesses *and* users.
Businesses can make very fine-grained choices around how much control they keep vs give to users.
But whatever rules they choose, that's what the rules are. Everything is onchain, transparent, and auditable by third parties.
You can make a very closed system where there's no escape hatch, and where the operator can edit the state root, but then users will know exactly what they are getting into, and @l2beat and armies of independent internet sleuths will help them.
You can also make a fully open system.
And you can also make a system where users know they can send a transaction on the L2 if they need to, but there's a large speed bump - the sequencer can throttle but not censor. This is effectively what Sony has done.
This is all free market at play. The main thing we need to watch for as an ecosystem is to ensure that
(i) a critical mass of financial activity happens in fully open environments
(ii) enough tools are available for users to understand the properties of the onchain environments they are spending their time in
You cannot legitimately argue that rollups are different "countries" than Ethereum.
This was strictly demonstrated just today with the sonieum censorship stuff.
Rollups are bound to Ethereum because you can change the state on an L2 from Eth mainnet.
L2s are like cities/states built on top of Ethereum.
Apps are like business within these economic zones.

































