tools/visit_webpage_tool.py

from langchain.tools import tool
import trafilatura

@tool
def visit_webpage(url: str) -> str:
    """Fetch and return the text content of a web page."""
    downloaded = trafilatura.fetch_url(url)
    if downloaded:
        text = trafilatura.extract(downloaded) or ""
        return text[:15000]
    return "Error: could not fetch page"