dama-cyber/distilled-novel-toolbox

This repository, written entirely in Chinese, is a reference library for authors of Chinese web novels (online serialized fiction published on platforms like Qidian, Fanqie, and Feilu). It is structured as a collection of skill modules, each covering a different part of the writing process, from early concept development through publication and platform monetization. The library has thirteen modules. Each module contains a core methodology document and a folder of deeper reference files on specific sub-topics. The modules cover areas including character design and ensemble casts, worldbuilding and power systems, pacing and chapter-level structure, genre selection and genre blending, emotional engagement and reader immersion, commercial strategy and platform-specific writing styles, compliance with platform content rules, and methods for editing AI-generated text to make it read more like human writing. There is also a module specifically about avoiding AI detection on platforms that flag AI-generated content. The README describes two primary creation workflows: one for short fiction in the 8,000 to 20,000 word range, and one for long serialized novels that may run to millions of words. The short-fiction workflow centers on a single emotional hook and one major reversal. The long-fiction workflow treats the project more like an engineering process, with daily word count targets, structured outlines, and a focus on distributing satisfying moments (described as "pleasure points") at regular intervals to maintain reader engagement. The recommended approach for AI-assisted writing is a mixed model: AI generates an initial draft covering roughly half the text, and the author then rewrites and refines the rest to inject personal style and remove patterns that platforms might flag as machine-generated. An example short story created using the library is included in the output directory. The project is licensed under MIT.