On November 21, a seemingly "normally authorized" transfer thrust the cross-chain protocol Chainflip into the spotlight: its Ethereum treasury was attacked on-chain, resulting in assets being withdrawn in batches. According to Onchain Lens and several media reports, the total amount stolen was approximately 4.86 million USD, including about 520.2 ETH, 3.25 million USDC, and 758,000 USDT. The most unusual detail is that the attack transaction was not triggered by common contract vulnerabilities but carried a valid signature generated by the treasury's aggregated signing key, completing the asset transfer "compliantly" in terms of permissions. This indicates that controlling the signing rights itself became the core breakthrough for the attack. Afterwards, the attacker began to disperse the funds, transferring some of the stolen ETH into Tornado Cash for mixing, further erasing on-chain traces. Why did such an attack occur on a cross-chain treasury claiming to use multi-signature or aggregated signature protection for large asset pools? What should the safety red line of cross-chain protocol treasuries be? These are the key questions this article attempts to clarify by examining this "legitimate signature" abnormal transaction.
4.86 million transferred with a legitimate signature: treasury lost in an instant
The attack on November 21 did not present an unusual contract invocation on-chain but instead a series of "seemingly normal" treasury withdrawal transactions. According to Onchain Lens's on-chain monitoring, the Chainflip Ethereum treasury address initiated multiple asset transfers in a short period: approximately 520.2 ETH, 3.25 million USDC, and 758,000 USDT were subsequently withdrawn, totaling about 4.86 million USD. The technical forms of these transactions are virtually indistinguishable from the withdrawals initiated for cross-chain settlements in daily operations—normal invocation paths, normal permission checks; the key point is that they all carried valid signatures authorized by the treasury's aggregated signing key, giving the external impression that "the treasury is transferring funds according to the rules," rather than "the treasury is being attacked."
The truly unusual aspect is the direction of the funds after they were withdrawn. Once assets flowed out of the treasury address, they did not remain in a single receiving address but were quickly split and dispersed to multiple wallets controlled by the attacker, making it challenging for simple monitoring on a single address dimension to reconstruct the complete fund path in real-time. According to a single source report, after the initial theft, the attacker further transferred about 2,909 ETH through multiple wallets, amounting to about 4.75 million USD. While this data awaits further cross-verification through more on-chain analysis, it is sufficient to indicate that the attacker deliberately increased the complexity of the fund flow. For any monitoring and tracking system that relies on "anomalous contract behavior" or "unauthorized signatures" to set risk control thresholds, such transactions that are formally compliant yet essentially represent unauthorized asset withdrawals constitute a direct and clear pressure test.
Not a contract vulnerability, but a breach of key permissions
In this theft of the Chainflip Ethereum treasury, what was truly lost on-chain was not the contract logic but the signing key controlling the treasury's deposit and withdrawal permissions. Traditional perceptions of attacks often hinge on smart contract vulnerabilities that rely on re-entrancy, price oracles, etc., to "bypass rules"; however, according to public transaction records, the key transfers used valid signatures generated by the treasury's aggregated signing key. The entire invocation path complied with the contract's established permission verification process, making it look more like a "normal treasury operation," except that the executor had transformed from a protocol party into an attacker.
To understand this difference, one must return to the commonly seen multi-signature or aggregated signature architecture in cross-chain treasuries: project parties do not use a single private key to control millions of dollars' worth of assets but instead set up a "multi-party consensus is required to unlock" lock. Only when a certain number of keys participate in the signature does the on-chain system recognize that the transfer has permission. Aggregated signatures compress multiple party signatures into a "qualified signature" at a cryptographic level; as long as threshold conditions are met, the contract will release the funds according to preset rules. If such aggregated signature private keys are controlled, the attacker does not need to alter any code or create any anomalous calls; they only need to construct formally compliant yet malicious signatures and transfer instructions, and the treasury will automatically execute them. This type of attack, which opens a hole at the permission level, exposes the long-ignored security shortcomings of cross-chain protocols.
The trust fracture in cross-chain treasuries: the red line in users' minds
For ordinary users, the core promise of cross-chain protocols is "give me the coins, and the system guarantees you can get them back at any time." The Ethereum treasury of protocols like Chainflip essentially assumes the role of cross-chain asset custody and settlement; once breached, it results not only in a loss of approximately 4.86 million USD worth of assets but also reveals cracks in the custody logic itself—users originally believed the treasury could only release ETH, USDC, and USDT according to preset rules, but now they realize that as long as the aggregated signing key falls into the wrong hands, "the rules" will effortlessly serve the attacker. For users who rely on cross-chain to move assets between different chains, a breach of treasury keys represents the most psychologically unbearable and hardest-to-forgive red line.
The deeper impact is that behind such aggregated signatures or multi-signature structures exists an underestimated premise: users do not solely trust code; they also implicitly place trust in the operational team's key management processes, the discipline of signing participants, and their security capabilities. In this attack, the related transactions fully complied with the authorization rules on-chain, the signatures were "valid," and the contract executed according to established logic. However, the final result was that treasury assets were withdrawn in batches into wallets controlled by the attacker. To external observers, such transactions are outwardly indistinguishable from daily operational instructions, and monitoring tools can only identify issues post-facto through anomalous amounts and paths. Users find it difficult to intuitively assess risks on-chain beforehand. If the treasury can complete a theft under the guise of "legitimate signatures," the safety boundaries that cross-chain protocols once established through technical architecture become blurred and unreliable at that moment.
Funds sent to Tornado Cash, on-chain tracking meets obstacles
After completing the initial transfer under the guise of "legitimate signatures," the attacker did not allow the funds to remain on exposed addresses but instead continued to split and funnel some of the stolen ETH into Tornado Cash. This on-chain action signifies that a complete path has formed from the Chainflip treasury to the attack address and then to the mixing contract, and all subsequent analyses can only anchor around these visible transfer transactions.
The working principle of mixing protocols like Tornado Cash is to pool multiple deposits from different users into a single fund pool and then anonymize withdrawals from new addresses at various times, without maintaining a public mapping of "one deposit necessarily corresponds to one withdrawal." For public chain analysts, what is clearly visible is that this portion of the ill-gotten gains entered the mixing pool, yet it becomes challenging to reconstruct precise paths for each ETH within the fund pool. While the link from the cross-chain treasury to the attack address remains clear, once it enters the mixing contract, tracking can only rely on statistical characteristics like amount distribution and time window to infer potential associations, rather than confirming ownership transaction by transaction as before. This directly raises the threshold for technical analysis and subsequent recovery efforts.
Challenges in cross-chain security revealed by the Chainflip incident
The attack on the Chainflip Ethereum treasury on November 21 does not stem from traditional contract vulnerabilities but rather from the key management and permission architecture of the aggregated signing treasury: the attack transaction used a valid signature generated by the treasury's aggregated signing key, meaning that once the threshold of keys is breached, the protocol itself will "legitimately" open the treasury for the attacker. This structural risk is amplified in cross-chain scenarios. Approximately 4.86 million USD worth of assets were stolen (including ETH, USDC, USDT, as reported by Onchain Lens through various media), but crucial details still contain information gaps: how exactly the private key was leaked or seized, whether the stolen USDC and USDT have been frozen or circulated again, the specific scope of user asset impacts, and whether Chainflip has put forth technical and asset disposal plans. These missing pieces of information directly determine whether the incident is categorized as an operational mistake, a key management disaster, or a more complex systemic failure. From an industry perspective, this case has become one of the negative examples of cross-chain treasuries, forcing security discussions to expand from "contract audits" to "key audits," "permission audits," and "process audits": follow-up attention should focus on whether cross-chain protocols will implement substantive adjustments in key custody methods, signing architectures (including threshold settings and multi-party participation mechanisms), risk disclosure, and emergency plans, as well as whether auditing parties and governance mechanisms can incorporate such key-level systemic risks into routine review and decision-making processes. These changes will determine whether cross-chain treasuries can still guard the last door on-chain when the next similar attack arrives.
Join our community, let’s discuss together and grow stronger!
Exclusive Hyperliquid benefits for AiCoin: https://app.hyperliquid.xyz/join/AICOIN88
Exclusive Aster benefits for AiCoin: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata
免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。




