Let me be sure I understand... editors send her papers to review, she sends them to you to do the actual work, then she returns them to the editors as her work. Is that correct? If so, you have two options.
You could, the next time you get such a paper, decline immediately and without making excuses. "I'm sorry, I cannot do this, and won't be able to do so in the future." If you're afraid of burning bridges, you might find a likely excuse, such as press of work.
The "nuclear option" is to accept the next paper, review it, and return it directly to the editor (assuming you can identify the publication and editor) with the notation that "Dr. X asked me to review this." You absolutely will burn bridges if you do that, but you won't get any more papers from her, either. Edit: In the comment below, Dan Romik is correct that what I suggested is passive-aggressive. However, if my summary in the first paragraph of the answer is correct, then the former supervisor is engaging in serial plagiarism. A report to the editor is equally nuclear, equally certain to burn bridges, and likely equally effective. (Before anyone says it's the same thing, I want to point out that responses to other current questions here take the position that reporting academic misconduct is not only appropriate, it is required. Of course, those answers apply to students... )
Edit: It was suggested in a comment that I add what happens if one burns bridges with one's former supervisor. That, of course, depends entirely on individual circumstances. The best generalization I can reach is that, the further along one is in one's career, the less damaging it is likely to be. If OP will depend on this supervisor for a recommendation following the post-doc, then it may be necessary to allow that former supervisor to continue to take advantage, or to fabricate excuses. If, at the other extreme, OP is the star of the lab and is certain of a tenure-track position at the same institution after the post-doc, then burning that bridge has less chance for harm.