Use LLM to generate and fix scraping code

[azhong-git]

Any thoughts on using LLM to write cheerio scraper code and fix the broken routes automatically?

(More context, I am author of https://github.com/lightfeed/lightfeed-extract - it is an end-to-end LLM extraction solution which might be not applicable here. But curious if there is a need to use LLM to create/fix scraping code to parse HTML to JSON)