I would argue that it would be better to perform the cyclization the other way. Using the two nucleophiles (OH and NH2) to attack primary halogens. The only issue is the building up that strange looking chloro beta-amino alcohol. I think this should work reasonable well.
I want to say hydroboration is the correct route, but I wonder if the chlorine is electronegative enough to stabilize the carbocation and cause the addition to go the other way and form the alpha-aminoalcohol...
I also put the dibromo, but really any dihalide should work. Or start with ethene, and add whatever halide you want...