ETH Liquidator
DRY-RUNAave V3 liquidation bot on Ethereum. Monitors 2500+ borrowers, flash-loan execution via Flashbots bundles. Zero-capital operation.
Key Numbers
At a Glance
2,500+
Positions Monitored
$0
Capital Required
Flashbots
MEV Protection
Atomic
Execution
Overview
About This Project
An Ethereum-native liquidation bot targeting undercollateralized positions on Aave V3, the largest lending protocol on Ethereum. The system continuously monitors 2,500+ borrower accounts, tracking health factors in real-time to detect liquidation opportunities within a single block of becoming eligible.
Execution is entirely zero-capital: flash loans provide the repayment asset, the liquidation captures discounted collateral, and the loan is repaid -- all within a single atomic transaction. This means the bot never holds inventory risk and can operate without upfront capital deployment.
All transactions are submitted through Flashbots private bundles, ensuring complete frontrunning protection. The bot competes in one of the most technically demanding MEV verticals on Ethereum, where milliseconds and gas optimization determine profitability.
Features
What It Does
Health Factor Monitoring
Tracks health factors for 2,500+ Aave V3 positions using optimized multicall batching. Detects liquidation eligibility within one block of threshold crossing.
Pair Ranking Algorithm
Scores liquidation candidates by expected profit considering collateral discount, gas cost, flash loan fees, and competition intensity -- prioritizing the most profitable opportunities.
Flash-Loan Execution
Zero-capital operation through atomic flash loans: borrow repayment asset, liquidate position, receive discounted collateral, repay loan -- all in one transaction.
Flashbots MEV Protection
Private bundle submission via Flashbots eliminates frontrunning risk and enables precise gas pricing to maximize profit on each liquidation.
Architecture
How It Works
Challenges
What Made This Hard
The Ethereum liquidation market is among the most competitive MEV verticals, dominated by professional searchers running custom infrastructure with co-located nodes. Achieving competitive latency requires optimized RPC pipelines and aggressive multicall batching. Gas price estimation for profitable bundles demands modeling the broader MEV auction landscape, since overbidding erodes profit while underbidding means losing to competitors.
Stack