Open the network tab when you initiate a cache prune in the new UI and you'll see a failed GET request to https://api.dagger.cloud//magicache/prune?async=true .
I'm not sure of the impact, but the // doesn't look right.
After initiating the prune, waiting a few minutes, hard-killing my local dagger engine, and starting it back up, I'm still seeing about 30s engine start time spent on downloading cache volumes - so I'm not sure if it was actually pruned.