This project offers a free, open-source AI-powered tool for coding interview preparation, targeting students and job seekers who find premium platforms too expensive. It provides core functionalities like problem analysis, solution generation, and debugging assistance using the user's own OpenAI API key, ensuring local processing and privacy.

How It Works

The application leverages Electron, React, and TypeScript with a Vite build process. It uses OpenAI's GPT-4o Vision API to analyze screenshots of coding problems and generate solutions with time/space complexity analysis. Users interact via a customizable, invisible window with global keyboard shortcuts for capturing problems, processing, and managing the interface. The architecture is designed for extensibility, allowing integration with other LLM providers.

Quick Start & Requirements

• Install: Clone the repository (git clone https://github.com/greeneu/interview-coder-withoupaywall-opensource.git), navigate into the directory, and run npm install.
• Prerequisites: Node.js (v16+), npm/bun, OpenAI API Key, Screen Recording Permission (macOS).
• Run: Execute stealth-run.bat (Windows) or chmod +x stealth-run.sh && ./stealth-run.sh (macOS/Linux). The window is invisible by default; use Ctrl+B (or Cmd+B) to toggle visibility.
• Docs: Contributing Guidelines

Highlighted Details

• Invisibility: Features a 99% undetectable window that bypasses most screen capture methods, though not Zoom versions 6.1.6+ or macOS native screen recording (Command+Shift+5).
• AI Capabilities: Utilizes GPT-4o for problem analysis, solution generation, and real-time debugging assistance.
• Customization: Easily adaptable for different AI models (Claude, Deepseek, Llama) and programming languages by modifying specific TypeScript files.
• Privacy: All processing and data, except for API calls, remain local to the user's machine.

Maintenance & Community

This is a community-driven, non-profit initiative. Support and feature requests are handled through GitHub issues. Contributions are strongly encouraged.

Licensing & Compatibility

Licensed under the GNU Affero General Public License v3.0 (AGPL-3.0). This is a strong copyleft license, requiring modifications to be shared under the same terms, even if run on a network server. Commercial use or linking with closed-source projects may be restricted due to the AGPL-3.0's requirements.

Limitations & Caveats

The invisibility feature has known limitations with newer Zoom versions and specific macOS screen recording tools. The project explicitly states it's not a full-service product and relies on community contributions for feature development, with no expectation of personalized support from the maintainer.