From what I can see, we do in fact pass the polarity information along but it looks like the Agilent-supplied executable fails to recognize it, and just claims everything as positive even when writing an all-negative list. Skyline, at any rate, is passing this kind of thing to our BuildAgilentMethod executable:
Compound Group,Compound Name,ISTD?,Precursor Ion,MS1 Res,Product Ion,MS2 Res,Dwell,Fragmentor,Collision Energy,Cell Accelerator Voltage,Polarity
Amino acids,Glycine.light,FALSE,76.039304,Unit,58.1,Unit,20,130,3.4,4,Positive
Amino acids,Glycine.light,FALSE,76.039304,Unit,53.5,Unit,20,130,3.4,4,Positive
Amino acids,L-Alanine.light,FALSE,90.054955,Unit,44.2,Unit,20,130,3.8,4,Positive
Amino acids,L-Alanine.light,FALSE,90.054955,Unit,29.3,Unit,20,130,3.8,4,Positive
Amino acids,L-Alanine.light,FALSE,90.054955,Unit,28.2,Unit,20,130,3.8,4,Positive
Amino acids,L-Alanine.light,FALSE,90.054955,Unit,27.3,Unit,20,130,3.8,4,Positive
Amino acids,L-Arginine.light,FALSE,173.1044,Unit,156.1,Unit,20,130,6.4,4,Negative
Amino acids,L-Arginine.light,FALSE,173.1044,Unit,131.1,Unit,20,130,6.4,4,Negative
Amino acids,L-Arginine.light,FALSE,173.1044,Unit,129.1,Unit,20,130,6.4,4,Negative
Amino acids,L-Arginine.light,FALSE,173.1044,Unit,85,Unit,20,130,6.4,4,Negative
(etc)