What problem does it solve? Turning a children's picture book into an interactive read-along practice site requires tedious manual work: transcribing pages, translating sentences, generating per-sentence audio, aligning word-level timing, and wiring up a playable web page. This Skill automates that entire pipeline while preserving the original story text, page order, and line breaks. ## Core Features & Use Cases - Transcription and Translation: Accepts page-by-page English text or page images, transcribes with visual verification, and produces one-to-one Traditional Chinese translations validated per page. - TTS Audio with Karaoke: Generates per-sentence English MP3s with word-boundary timing, optional consonant-to-vowel learning pauses for young readers, and atomic staging so failed runs never corrupt published audio. - Standalone Practice Page: Produces a file://-openable index.html with page tabs, play/pause per sentence, word-level highlighting across line breaks, mobile-friendly scrolling tabs, and optional Docker Compose deployment with Nginx. - Use Case: A parent photographs 12 pages of a child's English picture book; the Skill transcribes each page, translates every sentence into Traditional Chinese, generates audio, and delivers a verified bilingual shadowing site the child can use on a phone. ## Quick Start Build a bilingual read-along storybook site from these page-by-page story images, with English audio, Chinese translations, and word-level karaoke highlighting.
