Extract clean readable webpage text from a public URL, removing boilerplate for LLM-ready research or RAG ingestion.