PlugWiki

story-long-analyze Skill: Web Novel Teardown in Claude Code

Makes Claude take apart a long web novel chapter by chapter and write a teardown report, for authors who want to learn from a hit book.

Installs
17.6K
on skills.sh
Stars
8.2K
on GitHub
Files
17
besides SKILL.md
Updated
Nov 2026
last repo commit

story-long-analyze reads a novel you supply as a text file and writes a folder of analysis files. You get the golden three chapters, a quick preview, one summary per chapter, plot and rhythm notes, character and setting files, a relation chart, a style file and a main report called 拆文报告.md. The SKILL.md and its output are in Chinese.

It starts when you type /story-long-analyze or /长篇拆文, or say things like "拆这本书" or "深度拆解", or give a path to a novel file. It works in stages (index, first three chapters, batch extraction, plot and rhythm, characters and settings, report, style) and can resume after a break. It runs bundled Python scripts and needs Python 3 on your machine. It is one of 14 skills in the Oh Story set, and the README says to run /story-setup in your writing project after installing.

Good for

  • Web-novel authors studying a hit book before writing their own
  • Writers who need chapter summaries, plot units and a rhythm analysis of a long novel
  • Claude Code or Codex users who read Chinese and can run Python 3

Install story-long-analyze

Claude Code

  1. 1

    Copy this prompt.

    Install the story-long-analyze skill in Claude Code. Run this command:
    
    npx skills add https://github.com/zenstory-ai/oh-story-claudecode --skill story-long-analyze
    
    Then check that it loaded and tell me how to use it.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

From its listing on skills.sh. Source.

What's in the skill

When it loads

长篇网文拆文。保留黄金三章、逐章摘要、剧情、情绪、节奏、角色、设定和文风接口,以连续章节块完成因果、双时间线、关系与三维节奏分析;兼容旧成果直接使用、按需增强和断点续跑。含可选三层灵感库管道(灵感库、跨书灵感聚合、更新灵感库)。触发方式:/story-long-analyze、/长篇拆文、「帮我拆这本书」「拆这本书」「分析黄金三章」「深度拆解」「完整拆解」或提供小说文本文件路径。

Sections in SKILL.md

  1. 分析边界
  2. 对作者说话
  3. 运行约定
  4. 按时刻读
  5. Phase 1:确认对象并检查目录
  6. Phase 2:按检查结果走哪条路
  7. 固定交付接口
  8. 各阶段要点
  9. 状态与旧项目

Key facts

Installs on skills.sh
17,620 (Nov 2, 2026)
Last commit
2026-11-01

What it includes

  • Starts a teardown of a long novel from a text file or an existing analysis folder

    Also triggered by /长篇拆文 or phrases like "拆这本书". Asks for the book title, platform and text file path if you did not give them.

    Try: “/story-long-analyze 帮我拆这本书,原文在 ~/novels/盘龙.txt”

    Write · /story-long-analyze

  • Runs inspect_existing_assets.py to see what is already analysed

    Checks an existing analysis folder and picks a path: new analysis, continue a partial one, use finished results as they are, or repair missing stages.

    Runs code · Inspect existing results

  • Runs build_chapter_index.py to write chapter_index.csv

    Saves your text as 原文/原文.txt (converted to UTF-8 if it is GBK), then lists chapter boundaries and a hash for each chapter.

    Runs code · Stage 0: chapter index

  • Deep breakdown of chapters 1 to 3 plus a quick preview

    Writes one deep-dive file per opening chapter and 快速预览.md, then stops to ask you before the long run, unless you asked for it all at once.

    Write · Stage 1: golden three chapters

  • Hands chapter batches to chapter-extractor helper agents, 3 batches per round

    Each batch is read once and produces chapter summaries. If helper agents are not available it falls back to doing the work in the main session.

    Write · Stage 2: batch extraction

  • Plot units, storyline, rhythm and reader-emotion modules

    Works from the saved summaries without re-reading the book. Writes the 剧情/ folder.

    Write · Stage 3: plot and rhythm

  • Character files, setting files and a relation chart

    Writes 角色/ and 设定/, then render_relation_chart.py draws the chart from 角色/角色关系.md.

    Runs code · Stage 4: characters and settings

  • Writes 拆文报告.md, the file to read first

    Also fills in matching items in a 选题决策.md left by story-long-scan, if one exists.

    Write · Stage 5: main report

  • Writes 文风.md from quotes taken from the book

    Needs verifiable original sentences; if it cannot get them it fails instead of guessing.

    Write · Stage 6: style profile

  • Three-layer idea library; only writes under 灵感库/

    Runs only when you ask for 灵感库, cross-book aggregation or an update. Uses inspiration_index.py.

    Runs code · Inspiration library (optional)

  • Plan, commit, split, repair and mark stages in _progress.md

    The progress keeper. Its subcommands are plan, commit, split, repair-progress, mark-stage and digest.

    Runs code · manage_analysis_run.py

From the official docs

Permissions

Low risk

The skill runs bundled Python scripts and writes its analysis files into a folder in your project. Run it in a project folder you can review.

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • It runs Python scripts and writes many files

    Five bundled scripts index chapters, track progress and draw the relation chart. Run it inside a project folder you can review, and keep your original novel file elsewhere. github.com

  • Setup step adds hooks and agents to your project

    The README says /story-setup deploys hooks, helper agents and references for your agent. Read what it adds before you run it. github.com

  • Redoing a chapter means deleting its files yourself

    SKILL.md says to delete the chapter summary and the batch cache that covers it, then run again. github.com

Questions

What does story-long-analyze do?

It makes the assistant break a long web novel into files you can study. You get chapter summaries, plot units, rhythm and emotion notes, characters, settings, a relation chart, a style file and one main report. Everything is written in Chinese, in a folder named 拆文库/<book title>.

How do I install story-long-analyze?

Use the install box above for this skill alone. The README installs the whole Oh Story set with npx skills add zenstory-ai/oh-story-claudecode -y -g, then asks you to run /story-setup in your writing project and start a new session. /story-setup deploys the hooks and helper agents for your agent.

Does it work in Claude Code, Codex and the Claude app?

The README names Claude Code, Codex CLI, Google Antigravity, OpenCode, ZCode, OpenClaw and Reasonix. It does not document the Claude app. In Codex you call the setup skill as $story-setup.

Do I need an API key?

SKILL.md names none. The bundled Python scripts read only three environment variables: STORY_ANALYZE_CHART_FONT (a font path for the chart) and two STORY_ANALYZE_FAIL_AFTER_* variables.

Can it analyse a short story?

It is built for long novels. If your text is under 20,000 characters or it cannot find chapter titles, it suggests /story-short-analyze instead. If you insist, it still runs the long-form teardown.

Does it copy the book I feed it?

SKILL.md says it only carries over abstract mechanisms. It should not copy the original setting, character mix, key event chain, signature scenes or sentences.

Is it safe to run?

It runs Python scripts on your machine and writes many files into a folder in your project, so run it in a project folder you can inspect. The scripts write analysis files and a progress file. To redo a chapter you delete its summary file by hand, as SKILL.md instructs. Keep a copy of your novel text elsewhere.

Where does it fit with the other Oh Story skills?

The README lists the long-novel line as /story-long-scan (charts), /story-long-analyze (teardown) and /story-long-write (writing). The short-story line is the matching short-scan, short-analyze and short-write skills.

Information

updated Nov 2, 2026

Folder
skills/story-long-analyze

zenstory-ai/oh-story-claudecodeskills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 7
  1. story-long-analyze SKILL.md · github.com
  2. oh-story-claudecode README · github.com
  3. skills.sh listing · skills.sh
  4. zenstory-ai/oh-story-claudecode · github.com
  5. pipeline-ops.md · github.com
  6. manage_analysis_run.py · github.com
  7. render_relation_chart.py · github.com

Updated Nov 2, 2026.