Er, not specifically designed for the rovers.
I was sort'a hoping for software that corrected for the parallax shifts introduced to the rover panoramas because rover cameras do not pivot through their nodal point.
If you look at, for instance, this close pan of the rover deck, you can see the problem in it's gory details.
I would guess that a general solution to the rover panorama problem would require constructing a 3D model of the panorama scene from pairs of images and then constructing the final panorama from a selected viewpoint ( average of the nodal points of the views? ) from the model.
In any event far more information would be required to solve this problem than is available for the raw JPGs ( camera pointing, rover orientation, time of exposure... )
And it would have been so easy to construct camera mountings so that at least one camera was a true pan cam.
As you might expect, I already know of the gigapan site and camera mount.
If I remember correctly, I downloaded a rover panorama a few years ago to test the gigapan site. If anyone has a particular panorama that they might like see presented this way, just tell me which one and I will download it to the site.