How oz841119/quiz-mint-ai works
Project type and purpose
• Quiz Mint AI is a full-stack web application built with Next.js (App Router) and TypeScript.
• Purpose: provide a user interface to configure and generate quizzes using AI models (OpenAI-compatible or custom providers). It uses a single monolithic codebase to serve both client and server logic.
File structure
Top-level files and folders reveal key layers:
• Dockerfile, docker-compose.yml – containerization/development environment
• next.config.ts – Next.js build configuration
• src/app – Next.js App Router pages and API routes (UI + backend endpoints)
• src/AIService – domain layer handling AI-provider integration (OpenAI or custom)
• src/components – reusable React components (UI library, dialogs, sidebars, cards)
• src/configs – static lists (models, languages, quiz types)
• src/contexts – React Contexts for state management (exams, storage)
• src/hooks – custom React hooks (e.g., mobile detection)
• src/lib – utility modules and storage implementation (IndexedDB)
• Environment files (.env.local) – secrets and base URLs
Additional insights from README
• Describes development flow (Docker vs. npm dev server)
• Configuration of AI models and providers via src/configs/models.ts and AIService
• API route at src/app/api/generate-quiz/route.ts connects frontend requests to AIService