The vernacular alphabet gets created automatically on importing the data. This is based on the setup of the writing system and on the words existing for those letters. Use Custom ICU rules to control the alphabetized headers. Learn more about this in the FLEx help under "Writing System Properties, Sorting tab". See also this document, FieldWorks Sorting With ICU Rules by Ken Zook.

For example, if you want the letters č and č̣ not to appear under a normal c, but under their own letter headers, you could add them as an exception like this:

If you go to the dictionary view in FLEx, you will now see that it created additional letter headers (screenshot is just displaying the letter header č, but č̣ also got created):