Epidemiologist, Scientists, Statisticians, Historians, Data engineers and Data scientists are working on finding descriptive models and theories to explain COVID-19 expansion phenomena or on building analytics predictive models for learning the apex of COVID-19 confirmed cases, recovered cases, and deaths evolution time series curves. In CRISP-DM life cycle, 75% of time is consumed only by data preparation phase causing lot of pressures and stress on scientists and data scientists building machine learning models. This paper aims to help reducing data preparation efforts by presenting detailed data preparation repository with shell and python scripts for formatting, normalising, and integrating Johns Hopkins University COVID-19 daily data via three normalisation user stories applying data preparation at lexical, syntactic & semantics and pragmatic levels, and four integration user stories through geographic, demographic, climatic, and distance based similarity dimensions, among others. This paper and related open source repository will help data engineers and data scientists aiming to deliver results in an agile analytics life cycle adapted to critical COVID-19 context.