Honestly, I was thinking about spinosad... A natural soil bacteria that is lethal to some insects but not others. I was wondering if there are similar bacteria which could be deadly to varroa mites but harmless to bees and their larva. Maybe this is one area where agentic AI could prove beneficial... However, from the reading I've been doing, I understand why treating for mites is so difficult. Once they are attached to a host, it's essentially too late and being able to infect the mites BEFORE they attach to a host is astronomically unlikely apparently.
I don't know why I'm so committed to trying to find a solution to this problem. I don't have bees and know very little about beekeeping but they are such a critical cog in the ecosystem of life that I can't help but try...