
On-chain analyst Wazz claims to have identified a coordinated operation linked to 53 memecoin launches on Robinhood Chain, which extracted at least $18.43 million between July 10 and September 21, by exploiting a loophole in the Pons V2 anti-sniping tax to concentrate the supply and immediately resell to investors.
AI-generated summary
Robinhood Chain uses Pons V2 as the primary launch pad for its memecoins, with an anti-sniping tax of up to 99% to discourage immediate buying by bots, but allowing creators to exempt up to 32 addresses from this tax.
Fifty-three launches, the same mechanism. Pseudonymous on-chain analyst Wazz claims to have linked 53 memecoins launched on Robinhood Chain to a coordinated operation that allegedly extracted at least $18.43 million between July 10 and September 21.
The system was based on dozens of wallets authorized to buy the majority of tokens as soon as they were launched, before reselling them to other investors. The profits then fed the wallets responsible for financing the next operation. The total amount remains an estimate by the analyst, but on-chain data makes it possible to reconstruct several of these sequences. Explanations.
Key Points
Wazz attributes 53 memecoin launches to an operation that reportedly mined at least $18.43 million
Wallets exempt from the Pons V2 anti-sniping tax accounted for up to 86% of the supply
The Block confirmed the mechanism on ten tokens and traced the financing linking DRAFT to DEED
A portion of the funds left Robinhood Chain before being converted into 231,000 DAI on Ethereum
Robinhood Chain: The anti-sniping tax circumvented by the creators
According to Wazz, almost every targeted launch saw 70% or more of its tokens purchased immediately by groups of 70 to 200 wallets. Most of these memecoins came through Pons V2, the main launch platform for Robinhood Chain.
Pons sells the new tokens according to an automatic price curve. To discourage bots, the protocol applies a tax of up to 99% on purchases made immediately after launch. This then disappears in a few seconds.
However, the documentation authorizes the creator to exempt up to 32 addresses, in particular to allow a team to distribute its first purchases between several wallets. This functionality constitutes the heart of the observed mechanism.
Out of nine launches carried out since the end of August, the creators have exempted between 15 and 25 wallets. A bulk transaction then purchased the tokens for all of these addresses, just one to three blocks after their creation. The sales curve was emptied and the token joined a Uniswap v4 pool.
At the end of this operation, the creator and the exempt wallets controlled between 82% and 86% of the supply. A tenth project, EQUITY, followed a comparable method: 31 exempt addresses, of which 21 bought in one second, for a final concentration of 65.7%.
Wazz connects 45 launches through transfers between their collection and funding wallets. Four others share the same signing addresses for their bulk transactions, while four use a common collector wallet. These connections support the hypothesis of a coordinated operation, without making it possible to identify those responsible.
From DRAFT to DEED: The profits from one token finance the next
The circuit connecting DRAFT to DEED illustrates this organization. On September 14, 98 wallets that held DRAFT sent a total of 179.88 ETH to the same address in three seconds. The funds then joined a wallet starting with 0x9d06.
A week later, this wallet transferred 50 ETH to a new address. Part of the sum fed 0xf268, which distributed 15.98 ETH between 50 wallets. These included the creator of DEED, the wallet responsible for the initial purchase, and the other 24 addresses exempt from the tax.
DEED was launched forty minutes later. The creator and exempt wallets then held 86% of the supply, before starting to sell one second after opening. The Block records 130.75 ETH from these sales, to which are added 69.06 ETH in fees withdrawn by the creator. The media thus arrives at 199.8 ETH collected, or approximately 535,000 dollars. Wazz estimates the net proceeds at 212.94 ETH after deducting the initial funding.
On September 24, approximately 86.5 ETH then passed through the Relay bridge to Ethereum. They were converted into some 231,000 DAI and then transferred to a new address. Unlike issuer-controlled stablecoins, remaining ETH cannot be directly frozen at the token level, although intermediaries can block reported addresses.
CRUMBS represents the largest extraction estimated by Wazz with 3.12 million dollars, ahead of LEGS at 2.9 million and PINK at 1.44 million. Pons and Robinhood did not respond to requests from The Block at the time of publication.
AI outlook — possibilities, not facts
Robinhood Chain and Pons V2 will announce an update to their launch mechanism to limit the number of addresses exempt from the anti-sniping tax
Likely · Within weeks
An official investigation will be launched by financial regulators or on-chain analytics platforms to trace funds to Ethereum and identify beneficiary wallets
Possible · Within months

Apple has released iOS 26.7.1 update to fix zero-day CVE-2026-86950 in CoreGraphics. The company SlowMist warns of attacks targeting crypto wallets, although Apple has not formally linked the vulnerability to these thefts.

NEAR Intents claims to have blocked more than $50 million linked to the Bitget hack and froze an additional $503,000 using its SHIELD filter before solvers intervened.

China's Ministry of State Security claims that the anonymity of cryptocurrencies is an illusion, explaining that all transactions leave traceable traces via fiat-crypto exchanges and IP addresses, and cites the Bitfinex affair as proof, while banning the use of Bitcoin on its territory despite tolerance of detention.

Four days after Bitget's $387.5 million theft, on-chain investigator ZachXBT claims Chinese intermediaries are laundering funds on behalf of suspected North Korean attackers, publicly asking for help on Discord and Telegram when their XRP-bitcoin swaps on THORChain fail, captures of support tickets and transactions show.

The UNCTAD statistics portal (UNCTADstat) underwent a massive automated collection of 16,000 requests between April and June 2026. IP addresses and identifiers suggest a link to OpenAI agents, although the site only contains public data.

Solana is testing Alpenglow, an overhaul of its consensus replacing TowerBFT and Proof of History. This protocol aims to reduce transaction finality from 12.8 seconds to 150 milliseconds, improving efficiency for payments and bridges.