When do we know for sure that X is a good predictor of Y?

Root MSE of the Linear Fit Model is very small

R square is 0.9

p-value of Hypothesis Test on the Slope is less than 0.05

Correlation between X and Y is 0.9