I have written a blog post about how to set up a custom annotation file for the visualization of several genomes via mVISTA. The post includes an example using the plastid genome of Arabidopsis thaliana (i.e., is accessible to everyone via NCBI GenBank).

You can find that post here: https://blogs.fu-berlin.de/gruenstaeudl/2020/10/12/buena-vista-con-mvista/

Hey man, follow exactly the istructions (http://genome.lbl.gov/vista/mvista/instructions.shtml). I suggest you to format your file like the examples they give, which is not gff. They say they accept it but not sure about which version (it can also be gff2, it's an old tool!). So my final advice is to stick to these examples, which worked for me.

For example:

< 106481 116661 gene1 
106481 106497 utr 
107983 108069 exon 
109884 110033 exon 
111865 112023 exon 

> 39424 42368 gene2 
39424 39820 exon 
41401 42368 exon

> 77817 81088 gene3
77817 78820 utr 
79538 80107 exon