PAF development was started already in 2006 by Daniel Cano Fernández and Isidro González Caballero (see CHEP 2006 contribution) as a way to profit from a large (at that time) farm available at IFCA that could be used for preparing the analysis of the comming LHC data by the analysis groups at IFCA and the University of Oviedo.
At the start of the LHC data taking, PAF was resumed by Ana Y. Rodríguez Marrero and Isidro González Caballero by re-implementing part of the code and by adding support for distributed PROOF Cluster tools (see CHEP 2010 and CHEP 2012 contributions)
In 2015 Javier Delgado Fernández and Isidro González Caballero a new version of PAF was released with extra functionality, more flexibility and completely re-implemented code.