Hi,
I started using prodigy today, I've got my corpus as a list of Hebrew sentences in a simple utf-8 text file, when I open the prodigy web page I see the sentence in reverse, words are correctly from right-to-left but the whole sentence is reversed.
for example the correct sentence:
בוקר טוב מירושלים
is shown like this:
מירושלים טוב בוקר
Hi! That's very strange – we haven't had any reports like this and people have been annotating RTL languages without problems 
Which browser are you using? Are you using blank:he as the tokenizer? And what happens if you set "writing_dir": "rtl" (or "writing_dir": "ltr" for comparison) explicitly in your prodigy.json?
"writing_dir": "rtl" solved the issue, thanks!