How do train and test accuracy are calculated when cross-validation is used?

I have a dataset which is not split into train and test parts. I decided to use cross-validation to measure the train and test accuracies. I know that cross-validation is used to tune hyperparameters. I want to ask 2 questions:

1. If k-fold is chosen, then does cross_val_score(..).mean() return the test accuracy?
2. Is it logical to use cross-validation to get the test accuracy since I don't use it for tuning?

Thanks for now.