Smart Wrong Notebook (智能错题本) is an AI-powered system designed to help students efficiently manage, analyze, and review their incorrect answers. It addresses the common challenge of consolidating learning from mistakes by automating the generation of explanations, knowledge tags, and practice questions. The system benefits students by providing a structured approach to identify weak areas and improve academic performance through intelligent review and targeted practice.

How It Works

This system employs AI models, specifically Google Gemini or OpenAI APIs, to automatically analyze uploaded questions. It extracts key information to generate detailed explanations, assigns relevant knowledge point tags, and creates similar practice exercises to reinforce learning. Users can flexibly configure and dynamically switch between AI providers directly through the web interface. The application supports managing multiple notebooks organized by subject, features a smart tagging system with customizability, and offers multi-dimensional filtering options for efficient retrieval of specific questions.

Quick Start & Requirements

Deployment can be achieved using Docker, with a recommended docker-compose.yml file for easier management, or by cloning the repository and running locally. Local setup requires Node.js (v18+) and npm. Essential configuration involves setting environment variables for authentication (NEXTAUTH_SECRET) and AI provider credentials (API keys, base URLs, model names for Google Gemini or OpenAI). Database initialization is handled via Prisma commands (npx prisma migrate dev, npx prisma db seed). The project also supports Progressive Web App (PWA) functionality, enabling installation on mobile devices for a native-like experience.

Highlighted Details

• AI provider and model configuration (Google Gemini, OpenAI, or compatible APIs) can be dynamically updated via the web UI without requiring a server restart.
• PWA support allows users to add the application to their mobile home screen for an immersive, full-screen experience on iOS and Android.
• Features a robust, standardized K12 tagging system, with recent refactoring to a database backend for more flexible tag management.
• Includes multi-user support with data isolation and an administrator backend for user management.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), or a public roadmap were found in the provided README text.

Licensing & Compatibility

The project is licensed under the MIT License. This license generally permits broad use, including commercial applications and modification, provided attribution is maintained.

Limitations & Caveats

The README does not explicitly detail known limitations or bugs. However, functionality is dependent on the availability and cost of external AI services (Google Gemini, OpenAI). Proper configuration of environment variables, especially AI API keys and secrets, is crucial for setup. Default administrator credentials (admin@localhost/123456) should be updated for production deployments.