Basic method: LLM quality control without HTML content comparison

[PaulaMerle]

The basic way of evaluating main body extraction is giving the LLM the extracted text and instructing it to evaluate if the text is clean, it is separated well and understandable.

Acceptance criteria:

• first check is made without adding any html from the webpage, just evaluating the content completeness
• If the result is negative, html is added and LLM completes the extracted content based on the original web page