In Toolbox, it is challenging to get homograph numbers right, yet they matter if we will be validating cross references or promoting subentries.
The script AdjustHm.pl can be used to validate homograph numbers.
The ReadMe at the GitHub site explains which files to download and how to use the script(s).
This script will verify that entries that need homograph numbers have them, assign new numbers to those that need them, and report on anomalies such as duplicate homograph numbers.
Note: there are alternative ways to handle this process. For instance, technically subentries are candidates for homograph numbers, just like main entries. However, if the database you are importing has records where a subentry might be a homograph of its parent entry, it might be helpful to take a customized approach to this process, in order to make the "check cross references" step work better.
It might be better to run AdjustHm.pl in a manner that doesn't assign homograph numbers to subentries.
Then run the Check Cross References, so it is only looking at main entries for matches.
Then later run AdjustHm.pl again, to get hm on all the subentries. (But is this even needed, since FLEx will add them? It helps to make sure that \hm 1 gets on the main entry and higher numbers get on the subentries--I don't think FLEx accomplishes that.)
[All of this needs to be thought about more carefully, and maybe revised.]
Sense numbers should also be validated.
snNotInSequence.pl [Link needed] is a Perl script that checks the sequence of sn numbering in an entry and reports sequence numbers that are not starting from 1 or not in consecutive sequence.
Even though during the process of import, FLEx will adjust the sense numbers to start from 1 and number the following senses in consecutive numbers, it doesn't look like FLEx is also verifying any pre-existing references to an adjusted sense number. [Test this on a small database.] So if there are references to those originally irregular sense numbers, it is necessary to manually adjust the irregular sense numbers and any references to them before import.