C2C-coding-coach is an AI-driven platform designed to help users truly learn programming concepts and prepare for technical interviews, rather than just memorizing solutions. It targets beginners and job seekers by acting as a patient, interactive tutor that guides users through problems with personalized, chapter-based lessons and progressive hints. The platform aims to improve learning efficiency and deep understanding of algorithms and problem-solving patterns.

How It Works

The core approach leverages Large Language Models (LLMs) to provide a guided learning experience. Instead of direct answers, C2C employs Socratic questioning and breaks down coding problems into 6-8 AI-generated chapters covering syntax, data structures, problem-solving strategies, and step-by-step implementation. Python code execution occurs directly within the browser using Pyodide, enabling a zero-setup, offline-capable experience in the desktop application. Server-Sent Events (SSE) facilitate real-time streaming of AI responses for teaching, hints, and chat. LLM API keys are securely managed via Fernet encryption and configured through the UI.

Quick Start & Requirements

• Desktop App (Recommended): Downloadable executables for macOS (Apple Silicon), Windows (x64), and Linux (x64). Zero dependencies, offline-ready. Requires user-provided LLM API Key. First-time launches may trigger OS security warnings due to unsigned binaries; specific bypass steps are provided for macOS and Windows. Intel Macs are not supported by the desktop app.
• Docker: Clone the repository, configure .env.example, and run docker compose up -d. Requires Docker and an LLM API Key.
• Local Development: Clone the repository, set up backend (pip install, uvicorn) and frontend (npm install, npm run dev). Requires Node.js ≥ 20, Python ≥ 3.12, and an LLM API Key.

Highlighted Details

• AI Chapter-by-Chapter Teaching: Generates custom, multi-chapter lessons for each problem, including Mermaid algorithm diagrams and interactive exercises.
• 4-Level Progressive Hints: Guides users from conceptual direction to specific code snippets without revealing the full solution.
• In-Browser Python Execution: Pyodide enables immediate, zero-configuration Python code execution and testing within the web environment.
• AI Problem Generation: Users can describe a problem in natural language, and the AI generates a complete LeetCode-style problem with test cases.
• Personalization: Adapts teaching style and content based on 64 combinations of user experience, learning goals, and preferred learning methods.
• Multi-LLM Support: Integrates with Anthropic, OpenAI, Gemini, Qwen, Doubao, and GLM, with API keys managed securely and configurable via the UI.

Maintenance & Community

The project includes contributing guidelines (CONTRIBUTING.md) and automated release workflows in .github/workflows, indicating active development. No specific community channels (e.g., Discord, Slack) or sponsorship details are mentioned in the provided README.

Licensing & Compatibility

The project is released under the MIT license, which is permissive for commercial use and integration into closed-source projects. The desktop application has specific OS and architecture support (macOS Apple Silicon, Windows x64, Linux x64), with Intel Macs requiring alternative installation methods.

Limitations & Caveats

The desktop application does not support Intel-based Macs. All operating systems may display initial security warnings for the executable due to the lack of a code signing certificate. Users must provide their own LLM API keys, which are essential for the platform's core functionality.