Communitygithub.com

lzcjsyr/Book2Video

Converts long books and documents into Chinese video account reading scripts; first extracts text and establishes a recoverable coverage ledger, then generates drafts, multiple revisions, and final raw JSON based on the configured body length after coverage self-check passes.

What is Book2Video?

This AI agent skill, called Book2Video, automates the process of converting lengthy documents (e.g., books, reports) into polished Chinese video reading scripts. It first extracts text, plans bash-based reading windows, and maintains a coverage ledger to ensure full document coverage. Once coverage is verified, it drafts content according to configurable length targets (minimum 3000 characters), then iteratively refines the output through structural and final revision stages. The skill supports multiple revision rounds, audits differences, and finally generates a raw JSON output with title, cover copy, quotes, and the script body. It is designed for platforms like Cursor or Claude Code, and follows a strict workflow: understanding, angle selection, drafting, revision, and packaging. The skill emphasizes self-checking, requiring that coverage, writing standards, and revision logs are validated before delivering the final JSON.

Works with~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/lzcjsyr/Book2Video/tree/HEAD/skills/step1/book-video-script_V1

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Long Document Direct Bulk Reading Strategy

This skill is a thin entry point that breaks the long document processing into several independently maintainable modules:

  • references/reading-strategy.md: Bash continuous reading with each window not exceeding 23,000 characters, coverage ledger format, output path, and failure conditions.
  • references/writing-standard.md: Specific script requirements for video account reading scripts and the final raw JSON field format.
  • references/revision-workflow.md: Minimal but reviewable workflow for draft, structural, and final versions, including difference audit.

Body Length Configuration

This is the single source of truth for the body draft length in this skill. Other references must not duplicate specific word counts but can only reference this section.

draft_min_chars: 3000
final_target_chars: 3200
  • Before saving _draft_v1.txt, it must reach draft_min_chars.
  • All subsequent body drafts must not fall below draft_min_chars in principle.
  • The target length for content in the final raw JSON follows final_target_chars; total_length records the actual character count of content.

Main Workflow

Do not output in one pass. Follow "Understand -> Set Angle -> Write Draft -> Multi-round Revision -> Package Output":

  1. Understand Material: First read references/reading-strategy.md. Extract text, devise a Bash character window plan, establish and update the coverage ledger.
  2. Coverage Gate: Do not write drafts, revisions, or final JSON until coverage check passes.
  3. Type and Angle: Silently determine the material type, extract core questions, an explanatory framework, and a sharp angle relevant to the audience.
  4. Draft Writing: Read references/writing-standard.md, generate _draft_v1.txt reaching draft_min_chars.
  5. Multi-round Revision: Read references/revision-workflow.md, save structural draft, final draft, and revision audit.
  6. Package Output: From _draft_final.txt, generate title, cover copy, quotes, intro, and final raw JSON; unless the user requests to see intermediate steps, output only the project-usable raw JSON, and ensure it is parseable by json.loads.

Self-Check Requirements

Before submitting the final JSON, silently confirm:

  • Coverage requirements have been met according to total text length segmentation, covering the beginning, middle, end, and main sections, and that coverage_check.passed=true in the ledger matches reality.
  • The script complies with writing-standard.md regarding opening, rhythm, style, prohibited expressions, and revision requirements.
  • Each revision round re-reads writing-standard.md to ensure no requirements are forgotten.
  • The revision process has been saved per round as per revision-workflow.md, the final version originates from _draft_final.txt, and the audit corresponds to the actual differences in each round.
  • The JSON conforms to the final JSON output contract in writing-standard.md.

Individual skills in this repo

This repo contains 6 individual skills — each has its own dedicated page.

lzcjsyr/Book2Video

将 PDF、EPUB、DOCX 等书籍或长文档一次性转成约 4 分钟的中文读书带货短视频口播稿,以情绪共鸣、观看留存和购买欲为核心,直接输出 Book2Video 后续流程可读取的 raw JSON。用户提到读书带货、图书推广、卖书口播、商品卡、走心书单视频、情绪型读书文案或要求生成可直接进入视频流水线的带货稿时使用。

lzcjsyr/Book2Video

将长书、长文档、PDF、EPUB、读书资料、思想经典、社科著作、商业书或研究型长文转成更有传播感的中文视频号/抖音读书口播稿,并输出项目可用的 raw JSON。只要用户提到“读书解说”“拆书稿”“视频号口播”“把这本书做成短视频”“长文档转视频脚本”“生成标题/封面/金句/raw JSON”,都应使用此 skill。

lzcjsyr/Book2Video

Cody的公众号长文写作skill。当用户需要撰写公众号文章、写稿子、续写文章、根据素材产出长文时使用。触发词包括但不限于:写文章、写稿子、帮我写、续写、扩写、公众号文章、长文、出稿、按我的风格写。即使用户只是说"帮我把这个写成文章"或"用我的风格写一下",只要上下文涉及内容创作和公众号输出,都应该触发。也适用于用户丢过来一个PDF、brief、新闻链接、语音转文字或任何素材说"帮我写篇文章"的场景。不要用于短内容(小红书帖子、推特、朋友圈)或纯标题摘要生成。

lzcjsyr/Book2Video

将中短篇专业报告、公司财报、券商研报、招股书、宏观经济分析及含有较多表格和数据的PDF文档,转化为高品质、数据精准的中文视频号/抖音口播解说稿。只要用户提到“研报”、“财报”、“商业分析报告”或需要“解读数据”、“把报告/PDF做成短视频”,务必使用此 Skill。

lzcjsyr/Book2Video

简化版的文稿生成技能。能够直接读取图书/文档,提取核心论点,生成高品质、口语化的视频号/抖音读书解说文稿并输出标准 JSON 格式。

lzcjsyr/Book2Video

The fallback workflow for authoring custom HyperFrames video compositions at any length or format — longer or multi-scene pieces, brand / sizzle reels, montages, title cards, static loops, and freeform compositions. Input- and length-agnostic. If a specialized workflow clearly fits the input — a marketed product, a website, a topic explainer, a GitHub PR, existing footage, a short motion graphic, or a Remotion port — prefer it (see /hyperframes); use this only as the general fallback when none fit.

Related Skills