
Maya Protocol lost $1.7M after a hacker exploited six bugs in a single transaction. The CACAO token crashed 89% as the exploit mirrored THORChain's May hack.
Maya Protocol lost roughly $1.7 million on Aug. 18 after a hacker exploited six bugs in a single transaction, blockchain security firm CertiK said.
The attacker tricked the protocol into crediting a false balance of about 49 million CACAO tokens, even though the underlying pool held only around 168,000 CACAO. That inflated balance remained after a failed transfer. The attacker deposited a small amount, gaining more than 99% of the distorted pool. He then withdrew 48.87 million CACAO and 98.82 LINK, swapping them for Bitcoin and other assets.
The direct loss was $1.7 million, but developer Vini Barbosa said the total impact reached roughly $11 million after CACAO's price collapsed. The token fell from $0.115 to $0.013, a drop of nearly 89%.
Maya co-founder Aaluxx acknowledged the damage and said the team would “work to fix and recover in full.” He later told THORChain’s community podcast that the exploit relied on three older bugs that were harmless individually but dangerous when chained together. The same latent flaw existed on Maya, he said.
Maya is a friendly fork of THORChain, which lost about $10.7 million in May from a similar attack. THORChain’s post-mortem identified the same underlying bugs.
Aaluxx warned that AI tools let small teams probe codebases from more angles, helping defenders but also giving attackers new ways to find vulnerabilities. He said Maya and THORChain chose to stay independent rather than integrate, so Maya could keep running swaps while THORChain paused for weeks.
The incident comes as crypto hacks hit a record pace. TRM Labs reported 207 cryptocurrency hacks in the first half of 2026, the highest ever for a six-month period, with smart contracts increasingly targeted through multi-step attacks.
Aaluxx said Maya plans to accelerate development of Aztec Chain, an omnichain DeFi project drawing lessons from Maya, THORChain, and Rujira.
Prepared with AlphaScala editorial tooling from the source reporting linked above. Indexable analysis may include a cited Alpha Score value. Publishing checks screen each story before release. Educational coverage, not personalized advice.