See this post below but it doesn't give a final solution to your question.

It shows a direction to continue your project.

Prokka bacteria genome annotation

This second post shows, that people do use Prokka, but they have some doubts about.

Bacterial genome annotation

And the post suggests some alternative.

The third post provides another way.

Is there a way to run RAST annotation locally?

You have to choose.

Find the ORFs in your transcripts (e.g. with transdecoder) and then blastx (or translate and blastp) against a bacterial protein database. Transcripts without ORFs and those without hits you could process with RNA finding tools (tRNAscan, Infernal, SortmeRNA etc.) and blastn against ncRNA databases. No need to do gene finding.

In theory, I can’t see any reason this wouldn’t work. The annotations may not be fantastic however if it’s quite a broken up set of sequences etc.