I want to find a criterion for sheaves of modules to be injective. It would be great if one can such a criterion for sheaves of modules over a ringed space. But an answer for sheaves of abelian groups is ok.
I can prove that an injective sheaf is flabby and module of sections over any open set is injective. Is it sufficient to be injective? Can you formulate and prove some properties of injective sheaves, which do not follow from these two.