That is what layers and masks are for.
Collision bodies, such as your KinematicBody and your StaticBody, may be in one or more layers determined by the Layers property and set what layers they look in for colliders with the Mask property. Body A detects a collision with Body B if B is in a layer that's set in A's mask.
To make your KinematicBody pass through your StaticBody you'd set their layers and masks to make sure none of the StaticBody's layers match the KinematicBody's mask and none of the KinematicBody's layers match the StaticBody's mask.
This question is about the difference between layers and masks.