Good question. A simple way to do this is to use the following formula:

Square of Std error of forecast = sum of squares of the standard error of the fitted model and the standard error of each of the fitted model's parameters.

The std error of the model is the std dev of the model's residuals. statsmodels will give you the model's residuals in the results objects (e.g. RegressionResults.resid).

The std error of each one of the model's parameters is also usually available from the regression library you use. If you use statsmodels, you can get the std error of the params in the results object. (e.g. RegressionResults.bse).

Once you have all the standard errors, take the sum of their squares, then take the square root to get the standard error of the forecast.

Good luck!