No, it's not. It's false advertising and lifting an unsafe that high in the call stack is dangerous and a sign of poor code quality.
The point of advertising no unsafe in the readme is so that people do not have to worry if the author handled unsafe correctly.