It might be a variant of "No true scotsman" or we could call it an "empirical judgment" fallacy.
Why I suggest the former is this: What she's suggesting is a reverse of the No True Scotsman. The No True Scotsman fallacy is to do the following:
A: No scotsman would drink vodka.
B: McTaggart is a scotsman and he loves vodka [empirical defeater]
A: No true scotsman would drink vodka, so McTaggart is not a true scotsman even if blah blah.
What you're describing is the inverse:
A: I'm an honest person
B: well, you did tell lies on occasions P, Q, and R. [empirical defeater]
A: I'm an honest person ergo I don't believe you
What they share is the following:
- Empirical defeater
- Category claim about an individual
Where they differ is that one regards the negation and the other assertion, and one is first-personal and the other third.
The second way is just to say that what she's saying does not square with reality. I don't think that shows up on any list of fallacies, but it is fallacious none-the-less. I guess you could say what she says is valid but not sound and she's confusing validity with soundness. Or you could just say there's an empirical defeater for her argument.