菜单
Convert EPUB books into clean, narration-friendly plain text for local VibeVoice audiobook generation, with optional chunking for stable long-form TTS. Use when the user has an .epub and wants text cleanup, chunk prep, or a local audiobook workflow for VibeVoice.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | epub-to-vibevoice |
| description | Convert EPUB books into clean, narration-friendly plain text for local VibeVoice audiobook generation, with optional chunking for stable long-form TTS. Use when the user has an .epub and wants text cleanup, chunk prep, or a local audiobook workflow for VibeVoice. |
| tools | Bash |
| risk | low |
| source | local |
| date_added | 2026-05-02 |
Prepare EPUB books into pure, narration-friendly text for VibeVoice.
.epub and want local TTS/audiobook generation.cd ~/VibeVoice
source venv/bin/activate
python scripts/epub_to_vibevoice_text.py \
--input "/path/to/book.epub" \
--output "/path/to/book.clean.txt" \
--start_regex "^(Chapter|CHAPTER|Prologue|PROLOGUE|To you\.|When Red wins\.)"
cd ~/VibeVoice
source venv/bin/activate
python scripts/epub_to_vibevoice_text.py \
--input "/path/to/book.epub" \
--output "/path/to/book.clean.txt" \
--chunks_dir "/path/to/chunks" \
--chunk_chars 1800 \
--start_regex "^(Chapter|CHAPTER|Prologue|PROLOGUE|To you\.|When Red wins\.)"
cd ~/VibeVoice
source venv/bin/activate
python scripts/generate_audiobook_en.py \
--input "/path/to/book.clean.txt" \
--output "/path/to/book.wav" \
--voice en-soother_woman \
--cfg_scale 1.2 \
--device mps
chunk_chars: 1200–2200 is usually stable.resources/.Convert technical book chapters, papers, tutorials, or lecture notes into precise but listenable audiobook narration scripts, especially when the source contains LaTeX formulas and code in Lisp, C, Java, C#, or similar languages. Use this skill for math/code narration, SSML scripts, audiobook production plans, synchronized transcripts, and compact-vs-precise reading policies.
Help build, debug, refactor, test, and review ClojureDart applications that target Flutter, native mobile/desktop, web, or plain Dart. Use when the user asks about ClojureDart, .cljd files, cljd.build, deps.edn :cljd/opts, cljd.flutter, Dart package interop from Clojure syntax, Flutter widget construction in ClojureDart, hot reload, REPL, or ClojureDart testing.
使用捆绑的 Babashka 脚本修复 Emacs Lisp、Scheme、Common Lisp 等 Lisp 文件的括号/分隔符错误。 当用户提到 Lisp 括号错配、Paren Edit Death Loop、`.el/.lisp/.scm` 文件修复，或想在 Claude Code、Codex、Gemini 中批量修复非 Clojure Lisp 文件时使用。
从当前对话中提取 DDD 风格的通用语言术语表, 标记歧义并提出规范术语. 保存到 UBIQUITOUS_LANGUAGE.md. 当用户想要定义领域术语, 构建术语表, 强化术语, 创建通用语言, 提到 "通用语言", "术语表", "domain model" 或 "DDD" 时使用.
在 Clojure 项目中组合使用 `clj-nrepl-eval`、`clj-paren-repair-claude-hook` 和 `clj-paren-repair` 来完成 nREPL 求值与分隔符修复. 当用户提到 Clojure、nREPL、括号/分隔符错误、Paren Edit Death Loop、Claude hooks，或想在 Claude Code、Codex、Gemini 中验证和修复 `.clj/.cljs/.cljc/.bb` 文件时使用.
Classical Chinese concise communication mode. Responds in brief 文言文, preserving technical accuracy and modern terms when needed. Use only when user invokes /wenyan.