Just caught up on what happened with that Aave liquidation event back in March, and honestly the root cause is pretty wild. Chaos Labs dropped their analysis and basically confirmed it was a classic oracle configuration mess - the kind of thing that can cascade into serious problems if you're not watching closely.

So here's what went down: the CAPO risk oracle had this constraint issue where the snapshotRatio parameter was capped at only 3% growth every 3 days. It needed to jump from around 1.1572 to 1.2282, but could only creep up to 1.1919. Meanwhile, snapshotTimestamp just updated normally to a timestamp from 7 days back. That mismatch between the two parameters? That's where things fell apart.

Because of that misalignment, the exchange rate ceiling calculated by CAPO ended up about 2.85% lower than the actual market rate - roughly 1.1939 versus reality. That gap triggered the liquidation of around 10,938 wstETH. Not exactly small potatoes. The wild part is the protocol itself didn't take bad debt hits, so at least there's that.

What's interesting is how fast the response was. Chaos Labs and BGD Labs immediately jumped on it, tanked the wstETH borrowing cap down to 1 to stop further liquidation cascades, and manually realigned the parameters through Risk Steward. They recovered 141.5 ETH via BuilderNet to start compensating users, with the Aave treasury covering the rest. Total DAO compensation needed? Expected to be under 345 ETH.

This whole liquidation scenario is a solid reminder of how sensitive these oracle systems are. One configuration slip and suddenly you've got tens of millions in unexpected movements. Definitely something worth monitoring if you're holding positions in these protocols.