This commit introduces a new test suite for evaluating the HTML/CSS beautification capabilities of different AI models.
The suite includes:
- A baseline unstyled HTML file (`plain_html/index.html`) to be used as the common input.
- Standardized prompts for various design aesthetics (e.g., Professional, Minimalist).
- Initial results for the 'Professional' style from several models, including ChatGPT, Kimi, Gemini, Doubao, Deepseek,
and Yuanbao.