513516/amharic-book-writer-anygvt

# Amharic Book Generator A full-stack AI-powered application that generates complete, multi-chapter books in Amharic based on a title and description. It features precise length control, strict language adherence, and DOCX export. ## Features - **Language Strictness**: Guaranteed 100% Amharic output. Automatically detects and translates/removes English accidentally generated by the AI. - **Length Control**: Specify target page counts (10-300 pages). The AI plans chapters and word budgets to meet this target. - **Streaming Generation**: Real-time progress updates as the AI writes the outline, chapters, and polishes the text. - **History**: Autosaves every generated book. View, read, or delete past projects. - **Export**: Download any book as a formatted `.docx` file. ## Tech Stack - **Framework**: Next.js 14 (App Router) - **Language**: TypeScript - **AI Model**: Google Gemini Pro (via `@google/generative-ai`) - **Styling**: Vanilla CSS (Premium Dark Theme) - **Persistence**: Local JSON file (`data/history.json`) - Easy to deploy, no database setup required. - **Document Generation**: `docx` library ## Setup & Run 1. **Install Dependencies** ```bash npm install ``` 2. **Environment Setup** Create a `.env.local` file in the root directory and add your Google Gemini API Key: ```bash GOOGLE_API_KEY=AIzSy... ``` *Get a key at aistudio.google.com* 3. **Run Development Server** ```bash npm run dev ``` Open localhost 4. **Build for Production** ```bash npm run build npm start ``` ## Page Targeting Logic The app uses a deterministic word-count strategy to approximate page length: 1. **Definition**: 1 Page ≈ 300 Amharic words. 2. **Budgeting**: - Total Target Words = `Page Count` * 300. - Outline Stage: The AI creates a chapter plan. - Budget Per Chapter: `(Total Words * 0.9) / Chapter Count`. (10% reserved for Intro/Conclusion). 3. **Execution**: During generation, the prompting engine explicitly requests the specific word count for each chapter. 4. * …