Content analysis over a corpus: candidates from seed-phrase semantic search, keywords or the whole corpus, and an LLM classifier checked against human-labelled gold data, in a notebook or script.
This plugin declares no skills.