SeleniumScrapingTool only works the first time

[MaxPleaner]

Here's a minimum reproducible example:

from crewai_tools import (
    SeleniumScrapingTool,
)

scrape_website_tool = SeleniumScrapingTool()

url = "https://halfrottengoddess.bandcamp.com/album/live-recordings"
tool = SeleniumScrapingTool(website_url=url, css_element=".track_list .track_title, .track_list .track_time, .track_list .title a")

result = tool._run()
print(result)

result = tool._run()
print(result)

The first time tool._run() is called, it prints the desired info:

COMPLEX '22 - Demeter
COMPLEX '22 - Breach
COMPLEX '22 - Black Hole City
XX ATTACK FEST '23 - Whistle Heart
XX ATTACK FEST '23 - Girl in Space
XX ATTACK FEST '23 - Cryochrist
POINT EMERY SOLAR VAN '22 - Thane

The second time, it raises this error:

Traceback (most recent call last):
  File "/Users/maxpleaner/Desktop/AI_Agent/art_crew/tests/selenium_test.py", line 11, in <module>
    result = tool._run()
             ^^^^^^^^^^^
  File "/Users/maxpleaner/Desktop/AI_Agent/art_crew/.venv/lib/python3.12/site-packages/crewai_tools/tools/selenium_scraping_tool/selenium_scraping_tool.py", line 135, in _run
    self.driver.close()
  File "/Users/maxpleaner/Desktop/AI_Agent/art_crew/.venv/lib/python3.12/site-packages/selenium/webdriver/remote/webdriver.py", line 602, in close
    self.execute(Command.CLOSE)
  File "/Users/maxpleaner/Desktop/AI_Agent/art_crew/.venv/lib/python3.12/site-packages/selenium/webdriver/remote/webdriver.py", line 454, in execute
    self.error_handler.check_response(response)
  File "/Users/maxpleaner/Desktop/AI_Agent/art_crew/.venv/lib/python3.12/site-packages/selenium/webdriver/remote/errorhandler.py", line 232, in check_response
    raise exception_class(message, screen, stacktrace)
selenium.common.exceptions.InvalidSessionIdException: Message: invalid session id; For documentation on this error, please visit: https://www.selenium.dev/documentation/webdriver/troubleshooting/errors#invalidsessionidexception
Stacktrace:
0   chromedriver                        0x0000000101364e6c cxxbridge1$str$ptr + 2722840
1   chromedriver                        0x000000010135cd74 cxxbridge1$str$ptr + 2689824
2   chromedriver                        0x0000000100eae260 cxxbridge1$string$len + 90252
3   chromedriver                        0x0000000100ee8cf8 cxxbridge1$string$len + 330532
4   chromedriver                        0x0000000100f112e4 cxxbridge1$string$len + 495888
5   chromedriver                        0x0000000100f105ec cxxbridge1$string$len + 492568
6   chromedriver                        0x0000000100e7d3d8 chromedriver + 87000
7   chromedriver                        0x0000000101327f88 cxxbridge1$str$ptr + 2473268
8   chromedriver                        0x000000010132b1f4 cxxbridge1$str$ptr + 2486176
9   chromedriver                        0x00000001013099d0 cxxbridge1$str$ptr + 2348924
10  chromedriver                        0x000000010132bab0 cxxbridge1$str$ptr + 2488412
11  chromedriver                        0x00000001012faa60 cxxbridge1$str$ptr + 2287628
12  chromedriver                        0x0000000100e7b664 chromedriver + 79460
13  dyld                                0x000000018d90ab98 start + 6076

This is a problem for usage within Agents, because frequently they need to use the scraper multiple times.

Reinitializing the tool works, though:

from crewai_tools import (
    SeleniumScrapingTool,
)

scrape_website_tool = SeleniumScrapingTool()

url = "https://halfrottengoddess.bandcamp.com/album/live-recordings"
tool = SeleniumScrapingTool(website_url=url, css_element=".track_list .track_title, .track_list .track_time, .track_list .title a")
result = tool._run()
print(result)

tool = SeleniumScrapingTool(website_url=url, css_element=".track_list .track_title, .track_list .track_time, .track_list .title a")
result = tool._run()
print(result)

So for usage with Agents, I think I need to make a custom wrapper Tool which re-initializes the Scraper each time.