Code Tutor 💻

Code Tutor is a web-based IDE for learning and practicing Java, featuring an integrated AI Tutor.

Features

• Editor: Provides a code editor with Java syntax highlighting and IntelliSense.
• AI Chat Tutor: The tutor engages in conversation, applying the principle of minimal help by utilizing multiple AI agents to provide only necessary guidance.
• Language Server Protocol (LSP): Offers real-time IntelliSense and autocompletions.
• Code Execution: Executes Java test cases.

Tech Stack

• Frontend: Next.js (TypeScript), TailwindCSS
• Editor: Monaco Editor
• Language Server: Eclipse JDT LS
• Code Execution: Uses Piston
• AI: Supports Google Gemini and OpenAI LLM models

Wiki

An overview of the systems architecture and more can be found in the Project Wiki.

Getting Started

Prerequisites

• npm
• Docker Desktop

Setup

1. Clone the repository:

   git clone https://github.com/LucasWolke/code-tutor.git
   cd code-tutor

2. Install dependencies:

   npm install

3. Configure Environment Variables:

   Create a .env file in the project root and add your LLM API keys and backend URLS (only if not using local Docker containers):

   GEMINI_API_KEY=your_gemini_api_key_here
   OPENAI_API_KEY=your_openai_api_key_here
   PISTON_URL="https://emkc.org/api/v2/piston/execute"
   LSP_URL=your_language_server_url_here

4. Run the development environment:

   This command starts the frontend (Next.js) and backend services (LSP, Code Execution via Piston) concurrently.

   npm run start

   Send a POST request to http://localhost:2000/api/v2/packages with the desired language and version (e.g., {"language": "java", "version": "15.0.2"}) to install a runtime package.

5. Access the application:

   Open your browser and navigate to http://localhost:3000.

6. Stopping the services:

   To stop the backend Docker containers:

   npm run stop:services

   Use Ctrl+C in the terminal where npm run dev is running to stop the frontend.

License

This project is licensed under the MIT License.