Asymptomatic cases can spread the virus and there are well documented examples of this, although it appears to be much less common than transmission from people who have symptoms or who go on to develop symptoms in the few days after they test positive.
The only way to tell whether someone with no symptoms, who tests positive, is asymptomatic or pre-symptomatic is to wait and see whether they develop symptoms or not.
This virus is mostly spread through respiratory droplets. If someone has symptoms - is coughing and sneezing - they will spread a lot of droplets. But droplets can also be spread through singing, laughing and shouting, or even just through speaking and breathing normally.
So wear a face covering whenever you are close to others, especially indoors, because you might be infectious and not know. Even a cut up sock will catch a lot of your spit and snot droplets.
We don't yet know how likely children are to get infected or to pass it on. The vast majority of children get very mild symptoms or no symptoms which would suggest they are not shedding much virus, but for young children especially this must be balanced against their generally poor hygiene standards and their inability to socially distance.
We'd all love some certainty or hard and fast rules but there are none at this stage, just a careful balancing of risks, making allowances for a lot of unknowns, and different for each situation.
I don't think we can just say it's fine for kids to go back to school without a lot more data.
Nevertheless we need to make children's education, and all the other benefits they usually get through the school system, an absolute priority because there is no doubt that a huge number of children are greatly disadvantaged through not being at school. Education needs massive investment and it needs the best educational experts to come up with new ways of meeting our children's needs, which may not look anything like a traditional classroom but which also leave no children behind.
I don't have the answers but I can see a few 'known unknowns' we should be tackling.