Scan to download
BTC $77,393.84 +3.60%
ETH $2,422.71 +3.86%
BNB $645.29 +2.44%
XRP $1.48 +3.38%
SOL $89.02 +1.28%
TRX $0.3271 +0.14%
DOGE $0.0994 +1.64%
ADA $0.2587 +1.41%
BCH $455.69 +1.04%
LINK $9.61 +2.06%
HYPE $45.10 +3.93%
AAVE $115.92 +2.06%
SUI $0.9995 +1.16%
XLM $0.1743 +4.71%
ZEC $335.51 +0.40%
BTC $77,393.84 +3.60%
ETH $2,422.71 +3.86%
BNB $645.29 +2.44%
XRP $1.48 +3.38%
SOL $89.02 +1.28%
TRX $0.3271 +0.14%
DOGE $0.0994 +1.64%
ADA $0.2587 +1.41%
BCH $455.69 +1.04%
LINK $9.61 +2.06%
HYPE $45.10 +3.93%
AAVE $115.92 +2.06%
SUI $0.9995 +1.16%
XLM $0.1743 +4.71%
ZEC $335.51 +0.40%

law

Rhea Finance disclosed the reason for the attack, a flaw in the slippage protection logic led to a loss of 18.4 million dollars

According to RHEA Finance's official disclosure, the NEAR ecosystem lending protocol RHEA Finance (formerly known as Burrow Finance) experienced a margin trading feature hack, resulting in a loss of approximately $18.4 million.The attacker began laying the groundwork several days prior by creating multiple fake token pools on Ref Finance and injecting liquidity, constructing a malicious exchange route that exploited a vulnerability in the protocol's slippage protection mechanism—this mechanism did not account for the scenario where intermediate tokens were reused when calculating the minimum output of multi-step exchanges—leading to the borrowed debt tokens being directed into fake token pools controlled by the attacker, triggering a large-scale forced liquidation that ultimately drained the protocol's reserve pool. During the attack, the attacker deleted a total of 55 intermediate accounts to cover their tracks. Currently, the attacker has returned approximately 3.359 million USDC and 1.564 million NEAR to the RHEA lending contract, while another 4.34 million USDT has been frozen (of which Tether froze 3.291 million and NEAR Intents froze 1.053 million). The protocol contract has been suspended, and the team is collaborating with centralized exchanges for joint tracking and has notified relevant law enforcement agencies.

DGrid AI launches the AI agent platform DClaw, helping users build their personal local AI assistant with one click

AI infrastructure service provider DGrid officially launched its core product DClaw ------ a one-click personal AI agent deployment platform specifically designed for the DGrid ecosystem, aimed at significantly lowering the participation threshold for the open agent economy, providing out-of-the-box personal local AI assistants for individuals, developers, teams, and various communities.According to the official introduction, compared to the open-source framework OpenClaw, DClaw has achieved a comprehensive product upgrade. It focuses on true one-click agent deployment capabilities, compressing the technical configuration process that originally took hours into minutes; it natively integrates DGrid's unified model access API, allowing immediate access to top global models such as GPT-5.4, Claude Opus 4.6, Kimi K2.5 without the need for additional API key configuration; at the same time, it natively adapts to mainstream office and communication platforms such as WeChat, WeChat Work, DingTalk, and Telegram, enabling a single agent to operate uniformly across different environments.In addition, DClaw features a user-controlled persistent memory system and a hot-swappable modular skill plugin ecosystem, supporting multi-agent collaborative work to meet automation execution needs across various scenarios. Each DClaw instance can serve as an intelligent node in the DGrid network, achieving a deep integration of personal usage value and open ecosystem construction.Currently, DClaw has officially launched, allowing users to quickly build their own personal AI agents and participate in the construction of the DGrid open agent economy ecosystem.
app_icon
ChainCatcher Building the Web3 world with innovations.