This is a basic e-commerce website developed entirely using artificial intelligence tools on the Windsurf platform. No code was written manually; instead, instructions and prompts were used to guide the project development.
- Product Categorization: We use artificial intelligence to classify products into relevant categories.
- Product Search: AI helps users find products efficiently.
- Shopping Cart: Users can add and remove products from their shopping cart.
- Checkout: Users can complete the purchase process easily.
- Multilingual Support: The site now supports both English and Spanish languages using
next-intlfor seamless internationalization.
- Site Creation and Functionality: AI handled the website coding and implemented the required functionalities
- Repository Management: AI helped create the repository and manage version control, commenting on all changes made to the project. When requested, it initialized, versioned, and published the code on GitHub using command line tools, leveraging the GH CLI and my configured credentials.
- Documentation: AI handled the project documentation, including this README.md. I only made some content adjustments, as sometimes the AI tries to take all the credit in the text autocompletion 😃 and I helped too!
- Next.js: React framework for modern web applications.
- React: JavaScript library for building user interfaces.
- Tailwind CSS: CSS framework for rapid and responsive design.
- Framer Motion: Library for animations and gestures in React.
- Radix UI: Library of primitive and accessible UI components.
- React Hook Form: Library for form handling in React.
- Lucide React: Icon library for React.
- Zod: TypeScript schema validation library.
The project was created by Jorge Solano Kirk as a laboratory and proof of concept of what can be achieved without direct coding, using AI tools such as GPT-4o, Claude Sonnet, and Cascade Base. It serves as an experiment to explore collaboration with artificial intelligence in software development and to understand the limitations and benefits of these tools.
- Artificial intelligence cannot complete complex tasks, such as creating complex user interfaces or optimizing existing code. The project needs to be structured gradually as it develops, requesting incremental feature additions.
- Some errors were not identified when generating code, such as the shopping cart not updating when adding or removing products. We had to copy the console output or page content and paste it into the chat for the AI to address it.
- It was unable to understand that it should use the images I attached in the prompt for branding, attempting to generate its own SVG images. Only when the images were directly uploaded to the directory and specifically instructed to work with them was it able to make the adjustments.
This README file was also generated with artificial intelligence. The project demonstrates the potential of AI tools for software development as of November 2024.