In my opinion you really develop research skill in graduate school, but you can still do a lot in undergrade. The first key is build as solid of a foundation in physics as you can. I think some students make the mistake of ignoring fields that don't think they'll ever get involved with. After that you can build peripheral, but related skills: programming, electronics, machining, etc. Futher, it's a good idea to master the art of reading in your field. Even if you don't understand a lot of the jargon, at the undergraduate level, you can start with papers in Physics Today and Nature and work your way into more technical works.
One skill that often gets overlooked is writing. The craft of scientific writing is essential to learn if you plan on being a successful researcher because so many aspects of research are affected by one's ability to effectively communicate an idea - grant writing and getting papers accepted immediately come to mind. Along those lines, public speaking helps too.