Narrative planning supports the symbolic modelling of stories, but designing these models is logically complex and manually burdensome. To address this challenge, we introduce an integrated framework for extracting planning models from narrative texts. The pipeline iteratively synthesizes and refines planning models given narrative texts and both planning-specific and narrative-specific validation. Narrative models can provide insights into patterns across stories, and model generation automated by an LLM-powered framework makes this analysis feasible. This work demonstrates a preliminary implementation of our narrative planning model pipeline and a respective analysis as to how best to refine and improve our approach to our implementation as it matures. We found the automated extraction pipeline to have promising characteristics indicative of potential model viability in the future, given significant additional rigour throughout the model extraction and construction processes.