locking object to camera

My first thought in this project was to have a spacemouse directly manipulate an object, translate, rotate. This seems not possible, so I am considering a second approach.

I am considering a concept to lock an object to the camera. Then manipulate the scene with the spacemouse.

I would like to do some boolean operations that involve interactively fitting one set of shapes into other shapes.

Imagine fitting cube puzzle blocks together by having one block tied to the translation and rotation of the camera and the rest of the blocks in normal space.

Second analogy for a visual reference, the gun in a first person shooter game, stationary while the scene moves.

Is this a Grasshopper job?  I don't see the elements for acquiring camera translation and rotation. 

Many thanks for any ideas on approach.