PyWebFlx

Browser automation for already-open Chrome pages.

PyWebFlx is a Python library that connects to pages already open in Chrome via an extension (Manifest V3) + WebSocket. Unlike Selenium, it does not create a sandbox instance -- you automate the user's real browser.

---

Quick Start

import asyncio
from pywebflx import use_browser

async def main():
    async with use_browser(url="https://quotes.toscrape.com/") as browser:
        # View page structure
        structure = await browser.inspect(depth=5)
        print(structure)

        # Extract data
        quotes = await browser.extract_data(
            container="body",
            row=".quote",
            columns={"text": ".text", "author": ".author"}
        )
        for q in quotes:
            print(f"{q['author']}: {q['text'][:60]}...")

asyncio.run(main())

Why PyWebFlx?

	PyWebFlx	Selenium	Playwright
Uses already-open Chrome	:white_check_mark:	:x:	:x:
Reuses session/login	:white_check_mark:	:x:	:x:
Async/await API	:white_check_mark:	:x:	:white_check_mark:
No WebDriver/binary	:white_check_mark:	:x:	:x:
AI-optimized inspect	:white_check_mark:	:x:	:x:
Setup	Chrome Extension	WebDriver + config	playwright install

How it works

Python Script                   Chrome Extension
      |                                |
      |  1. Starts WebSocket :9819     |
      |<------------------------------>|  2. Connects automatically
      |                                |
      |  3. JSON command               |
      |------------------------------->|  4. Injects script into tab
      |                                |     via chrome.scripting
      |  5. JSON result                |
      |<-------------------------------|
      |                                |

Main features

• use_browser() -- Connects to tabs by URL or title, opens if not found
• click(), type_into() -- Element interaction (CSS, XPath, text)
• get_text(), extract_data() -- Structured data extraction
• inspect() -- Summarized DOM view, optimized for AIs
• extract_table() -- HTML table extraction with pagination
• wait_element(), element_exists() -- Synchronization
• execute_js() -- Custom JavaScript