GitHub Copilot CLI builds iOS app to track your mystery boxes—delivers functional MVP buried under debugging hell and dated SwiftUI vibes

Michelle Mashutian has developed BoxFinder, an iOS app that helps users track physical storage containers, as part of the GitHub Copilot CLI Challenge. The app allows users to create boxes with photos, add photos of contents, and manually tag items, enabling search by keywords to quickly find boxes and their locations. Mashutian utilized GitHub Copilot CLI to generate the initial code structure and iterated on the app using the command-line interface. The development process involved collaborating with the AI model to resolve issues, such as downgrading APIs to support iOS 16 and enabling photo taking functionality. Although the model struggled with UI refinement and debugging, Mashutian found the experience valuable in testing an AI-first iOS workflow, highlighting the importance of clear specs and careful direction. The app, while still in its early stages, addresses a real-world problem and demonstrates the potential of AI-assisted development, with implications for the future of software development and collaboration between humans and AI models.