A good way to get started is looking at one of the examples scripts.
This shell script calls different shimming-toolbox command lines functions to perform a whole shimming scenario. Acquisitions are downloaded from a Github repository and output text files and quality control figures are generated.
This function will generate static and dynamic (due to respiration) Gx, Gy, Gz components based on a fieldmap time series (magnitude and phase images) and respiratory trace information obtained from Siemens bellows. An additional multi-gradient echo (MGRE) magnitude image is used to generate an ROI and resample the static and dynamic Gx, Gy, Gz component maps to match the MGRE image. Lastly the average Gx, Gy, Gz values within the ROI are computed for each slice.