Axios

My project is a CRE based OEV liquidation pipeline for lending markets. CRE workflow is triggered immediately to a oracle price change event and it asks the indexer which borrowers position would fall below HF < 1 under that trigger price, without changing the lending market price yet. If liquidatable positions exist, CRE calls coordinator to open a public auction round where MEV-bots compete for liquidation rights and the highest bid wins. At reveal, the lending-market price is updated onchain and liquidation executes in the same transaction atomically. If the winning bidder fails, fallback starts: a watcher continuously tracks ERC-8004 AgentIdentityRegistry events and maintains a live directory of registered liquidator agents with endpoint, health, and reputation data. CRE then queries that directory and calls fallback agents one by one based on reputation until one completes the liquidation. This reduces failed liquidation rounds, improves execution reliability during volatile price moves, and helps prevent bad debt in the protocol in extreme scenario's.

In Future, I plan to integrate more risk features into the pipeline, like automating CRE workflow with ERC-8004 registered liquidity providers for extreme market scenarios. If supply or borrow moves cause heavy utilization or liquidity imbalances, the CRE workflow query the registry and call these providers to supply liquidity and help the market stabilize interest rates.

It is built with Chainlink CRE to react to price changes, a indexer which tracks borrower health, a coordinator which runs liquidation auctions for OEV capture, contracts to update price and execute liquidations onchain, ERC-8004 Identity Registry integration to discover fallback agents, and a live dashboard to monitor the entire workflow.