We have had a Bug Bounty program in place at Kraken for nearly ten years. This program is run internally and is fully staffed by some of the brightest minds in the community. Our program, like many others, has clear rules of the road…
1. Do not exploit more than you need to in order to prove the vulnerability.
2. Show your work (i.e. provide a proof of concept)
3. What you extract you return immediately
Since OZ has publicly disclosed the details of the problematic integration between two specific standards – ERC-2771 and Multicall – we now re-share our analysis of the related exploit.
The interesting aspect of this exploit stems from the way the public burn issue was taken advantage of, specifically the inconsistency between the Forwarder.execute() function and the TokenERC20.multicall() function. Notably, the attacker persistently targeted tokens vulnerable to this inconsistency.
Specifically, the Forwarder appends 'req.from' at the end of 'https://t.co/V3O55570QZ'. However, when the multicall() function processes the parameters, it excludes 'req.from' before executing the actual call. Consequently, when the burn() function invokes _msgSender(), it extracts 'req.from' from the end of the calldata for TrustedForwarder. Due to the earlier processing by multicall(), 'req.from' has already been discarded. As a result, the last 20 bits of 'https://t.co/V3O55570QZ', which can be manipulated by the attacker, are used.
Our analysis is based on this attack TX: https://t.co/gpON29xMtF
. @GROKERC20 was attacked due to the lack of the slippage control, and the loss is ~$56K (26 Ether).
Specifically, in the _transfer function of the #GROK token contract, the swapTokensForEth function can be triggered if the contract balance is greater than _taxSwapThreshold. By abusing this, the attacker could harvest more #GROK tokens with a reverse swap.
Finally, the attacker converted those #GROK tokens back into ETH.
https://t.co/5l6sBmLXWn
. @thestandard_io was exploited due to the lack of slippage protection when swapping out the collateral, and the loss is ~$290K.
The attacker first created a CDP (SmartVaultV2) with 10 WBTC as collateral to mint 290,000 EUROs. Subsequently, the attacker forced the SmartVaultV2 contract to execute a swap in the WBTC/PAXG pool, which was manipulated through the sole position the attacker had opened before. It allowed the attacker to drain liquidity and seize all WBTC collateral from the SmartVaultV2 contract, resulting in bad debt for the protocol and profits from the minted EUROs.
Attack Tx: https://t.co/nVnmZAErRv
MEV bot got exploited via unprotected callback for 8.88 ETH (~$16k) with a NFT called "Agent Origin (viet_nam)"
Lmao at the similarity to Agent Orange
0x0802cb621535999de2084b49f257d8bd0919cbfe03b186651689a74536e1a792
Due to the affected protocol being paused, here is a preliminary analysis of this @Platypusdefi incident:
1. It's a Flashloan attack, with a total loss of ~$2M. The Hacker manipulated 'cash' and 'liability' which affected the swap price.
2. The first attack tx is: https://t.co/esLyNWsPYc
3. There are 2 attackers who carried out the attack:
- 0x0cd4fd0eecd2c5ad24de7f17ae35f9db6ac51ee7
- 0x464073F659591507d9255B833D163ef1Af5ccc2C
4. One of the attackers made a mistake, and the Platypus team successfully rescued $575K! The rescued funds have been transferred to:
0x068e297e8ff74115c9e1c4b5b83b700fda5afdeb
5. We identified that the forked project @hummusdefi has the same issue and immediately contacted the Platypus team to see if they could be notified. hummus defi has been informed and just paused the protocol.
6. This incident continued with multiple attacks. Phalcon Block is an attack prevention system. You're welcome to join the waiting list.
2 weeks ago, @Balancer and its fork @beethoven_x sufferd attacks, with total losses ~$2.1M. The subtlety of the bug and its exploitation have remain under-analyzed. In this report we deliver a comprehensive community-engaging analysis.
https://t.co/3kOYltMtnD
Balancer is aware of an exploit related to the vulnerability below.
Mitigation procedures have drastically reduced risks, but are unable to pause affected pools.
To prevent further exploits, users must withdraw from affected LPs.
https://t.co/PDzX32gqeS
You either need to be an idiot or outright malicious to tweet out potential vulnerabilities @BlockSecTeam while there is an ongoing incident
Adding screenshots and asking the potential victim to DM you? IN A PUBLIC TWEET?
I wont even quote your tweet/s but what the actual fuck?