Narremgen: A Library for Generating Artificial Structured Texts Of Advice From Narrative, Emotional, And World-Context Templates.

Narremgen is a Python package for structured narrative text generation. It combines narrative structures (SN) with emotional dynamics (DE) to produce coherent short texts, which can be assembled into full booklets of advice or topic- or question-driven answers, optionally organized into chapters. The package includes about 50 SN templates, extended to roughly 110 for broader use cases, and defines 10 initial DE patterns. It also provides a reproducible, multi-batch pipeline for controlled corpus generation with large language models, using prompts and an iterative, multi-step process for each batch. The resulting corpora may be useful for education and learning, and for benchmarking, comparing, or training LLM outputs.