I looked through the code base to see if there was anyway to do this. Currently the pose is tied to the score table so they both get dumped to the PDB file by the job distributor. If you are looking to reduce file size, you could pass -out:pdb_gz to zip the output files before you parse them. Alternatively you could look into silent files that can greatly reduce the file size and can be manipulated by many downstream rosetta applications. Can you comment a little further on the end goal and I can help find a solution?
Sounds like a job for Rosetta tools rather than the apps. If you just want to do PDB manipulations and prepare them for Rosetta, checkout the tools directory that comes with Rosetta. Inside is two scripts for exactly what you are talking about. python tools/renumber_pdb.py and python tools/protein_tools/scripts/clean_pdb.py
Clean_pdb.py is really nice because it prepares the structure for Rosetta and can automatically pull it from the PDB if you don’t have the file.