Skyvern-AI / skyvern

Automate browser-based workflows with LLMs and Computer Vision
https://www.skyvern.com
GNU Affero General Public License v3.0
10.78k stars 739 forks source link

make sure we do a web scrape before data extraction at the end of task to ensure the refreshness of the scraped data #1152

Closed wintonzheng closed 3 weeks ago

wintonzheng commented 3 weeks ago

[!IMPORTANT] Ensure data scraping is refreshed with screenshots before data extraction in agent.py, handler.py, and scraper.py.

  • Behavior:
    • Ensure data scraping is refreshed before data extraction in agent.py and handler.py.
    • Modify refresh() in scraper.py to always include screenshots.
  • Prompts:
    • Update check-user-goal.j2 to include screenshots in user goal verification.
  • Functions:
    • Remove with_screenshot parameter from refresh() and scrape_website() in scraper.py.

This description was created by Ellipsis for f0f6d49e19e8ae26ff3efb1c87697c222858b44d. It will automatically update as commits are pushed.