Many times when we use on-chain applications, we rarely pay attention to "where the data comes from." But this is actually very critical. Because all execution is premised on input data. If the data has deviations or is unstable, even the best mechanism will struggle to guarantee results. What WINkLink is doing is to make this layer as stable as possible. By obtaining data from multiple nodes, then verifying and aggregating it, and finally outputting unified results.