Seeing Cage's brilliant scripts for vertex-level editing made me think about the following: what if the boxes generated could be made to track the vertex they were created from, and then be used to parent objects to? For example, a nose ring that stays in place even when the face is morphed. Or a spike attached to a chin that stays put when the mouth is opened. The script would need to create two boxes: one for the vertex, and another placed at a user-defined distance from the first, along the vertex's normal (for point-at purposes). My idea for how to generate the boxes is as follows: the user selects several faces surrounding a single vertex, using the grouping tool. This would make it easier to precisely choose which vertex gets used.