The tool features DOM Tree compression, automatic ID generation, conversation context, a Selenium wrapper, and training data. Users can set up the Selenium driver and execute prompts using the tool's wrapper. The tool also offers an interactive mode for real-time natural language commands for web application interaction. Users can fine-tune the AI model for more accurate results by converting HTML content to Markdown and training the model with the converted data. Contributions to the tool's development are welcome.
Key takeaways:
- The project introduces an AI-driven tool to automate website testing through Selenium, making web testing accessible and efficient.
- The tool has features like DOM Tree compression, Automatic ID generation, Conversation context, Selenium wrapper and Training data.
- The tool allows real-time natural language commands for web application interaction, and users can change URLs, navigate, click, input data, and exit as needed.
- The tool offers the possibility of fine-tuning the AI model to ensure more accurate results, by converting HTML content to Markdown and using the converted data for training.