Prosodic computational literacy is an important goal for students of acoustic phonetics, especially those from endangered language communities in less affluent countries. There are several convenient ‘off-the-shelf’ packages for prosody computation, including Praat, ProsodyPro, Prosogram, ProZed, Win-pitch, and many convenient Praat scripts. However, experiments typically require small hybrid intersections of functionalities of these packages together with spreadsheets, R, Praat scripting or Python. Python was chosen in order to enable non-hybrid, seamless embedding of small tools into larger systems for exploratory research, because of scalability, and because of the availability of extensive Python libraries to support in-depth in-sight into filters and transformations rather than using ready-made complex functionalities. A design criterion for the toolkit is overall coherence and clarity of structure. The tools cover the analysis of speech signal annotations, and a modulation-theoretic approach to the demodulation of speech signal amplitude modulation and frequency modulation. Comparison of results is enabled by provision of distance measurement and hierarchical clustering techniques. The approach has been evaluated in practice in a range of publications and in teaching.