[IsTom]

It might be using some nonlocal context e.g. something similar was proved earlier in the proof and the reader is assumed to remember and/or generalize this or there's some assumptions that are stated only in the beginning. There's also probably a bit of skipped proofs of the kind "I'll be able to do this if pressed, but so will the reviewer".