Militaries throughout history (a pretty broad statement, I know) generally are and were pretty big on written reports, ranging from fairly informal "we won"-type messages to complete reports meant to give superiors a good understanding of where the enemy was, in what numbers, and what happened. Beyond that, militaries were producing absolutely outrageous quantities of records related to pretty much all of their activities by the late 18th-century. Here are the finding aids for the War Department Collection of Revolutionary War Records; apparently a lot of the collection was destroyed over time, but you can see that in addition to reports (93.2.2) they saved all kinds of other stuff that can help historians verify reports and accounts of battles in letters, like recruiting and quartermaster reports.

At this point in history there's no real barrier to scholars from both belligerents viewing the Revolutionary War records of the other, either, which means that either side's account of a battle can be checked against the other's for accuracy. Between that, the "forensic" evidence of non-correspondence records, and the sheer volume of the accounts of battles available in most instances, these battles are well-studied enough that you can be reasonably sure that most battles are accurately described.

If you're suspicious of something, the best advice I can give is to start by Googling the author to look for controversy (N.B. - A controversial interpretation of events doesn't always mean the interpreter is wrong.) or Googling the topic and looking for accounts agreeing with the author.