Novel Outline turns a novel and a few target parameters into the first stage of an AI short-drama production. You give it the episode count, episode length and genre, and it produces an adaptation note, a cast table, a payoff-beat table, per-episode synopses and an asset list.
Fourteen quality gates run as code, not as a checklist for the model: character tier limits, a main-scene cap that scales with episode count, payoff spacing of three episodes or less, a required hook per episode and more. A checkup mode runs only the gates on an outline you already have. The skill is written in Chinese and its report defaults to Chinese, with an English option.
What You Get
- An outline.json file with adaptation notes, cast, beats, episodes and assets
- A Markdown version of the outline
- A single-page HTML report with KPI band, decisions, payoff timeline and quality gates
- A checkup mode that diagnoses an existing outline without rewriting it
Common Use Cases
- Adapting a web novel into a 60-episode short drama outline
- Checking an existing outline against payoff spacing and hook rules
- Preparing the upstream file that character, art and script stages read
What Makes It Unique
The pass or fail decisions are made by script, so a model cannot talk its way past a gate. Four of the five parts are written by the model and the fifth, the asset list, is computed from the episode data. A user approval step sits before the episodes are written, so a wrong direction costs one round of skeleton and not the whole script.
Under The Hood
The skill is an instruction file with Node.js scripts that use only the standard library. It runs shell commands to split a long book into volumes, validate the outline against the gates and render the report, and it starts sub-agents to summarise volumes and write batches of episodes. It reads the novel text file you point it at and writes JSON, Markdown and HTML files into an output folder. It does not fetch web pages or call outside APIs, and it needs no API key.