Libraries are usually for three dimensional data such as MSD data where you can extract a spectra. In the case of two dimensional data from a FID, the best you could do is to create a calibration table from known standards and use that for identification by retention time and concentration. This is not a library , but simple identification on your system using standards . I hope this helps.
Sorry I should have not use the word library. It is only for the internal purpose just to identify the peak and it is not for the quantification and it is only for the quality check purpose.