In the past few days, there have been two more pieces of Aave drama: one is a $27 million erroneous liquidation caused by a wstETH oracle error; the other is a user executing a collateral swap function through the Aave frontend, resulting in a $50 million loss due to slippage. I strongly agree with @lukecannon727's point in the first image: every project team needs to add an abnormal transaction circuit breaker feature to their frontend, just like payment software intercepts transactions when it detects an abnormal environment. It has been proven that relying solely on user confirmation is ineffective. At the same time, Aave's frequent dramas are pushing users towards its competitor Spark: in addition to Sun transferring money to Spark, some large holders have moved their wstETH to Spark after the wstETH oracle incident. (The Spark oracle reads the underlying asset prices from the issuer rather than the market trading prices, which poses less risk). It seems that in the eyes of large holders, Spark has already become the top alternative to Aave.