I will start with a brief review of Landscape and Swampland ideas in general and the different versions of the de Sitter Swampland conjecture in particular. This will naturally lead to the question how trustworthy our best models stringy de Sitter space, such as KKLT, are. The rest of the talk will then focus on recent progress in establishing KKLT more firmly by studying the underlying 10d geometry. This involves, in particular, understanding certain highly singular couplings between bulk and brane field in type IIB supergravity.