Selenium-Profiles

• Overwrite device metrics using Selenium
• Mobile and Desktop emulation
• Undetected by Google, Cloudflare, creep-js using selenium-driverless
• Modifying headers supported using Selenium-Interceptor or seleniumwire
• Touch Actions
• dynamic proxies with authentication
• making single POST, GET or other requests using driver.profiles.fetch(url) (syntax)
• headless unofficially supported
• apply profile on already running driver with driver.profiles.apply(profiles.Android())
• use of seleniumwire

Sponsors

This project is currently not being sponsored.

for the latest features, have a look at the dev branch

Getting Started

Dependencies

• Python >= 3.7
• Chrome-Browser installed

Installing

• Install Google-Chrome (or another chromium-based browser)
• pip install selenium-profiles

Start Driver

from selenium_profiles.webdriver import Chrome
from selenium_profiles.profiles import profiles
from selenium.webdriver.common.by import By  # locate elements
from seleniumwire import webdriver


profile = profiles.Windows() # or .Android
options = webdriver.ChromeOptions()
options.add_argument("--headless=new")
driver = Chrome(profile, options=options,
                uc_driver=False
                )

# get url
driver.get('https://abrahamjuliot.github.io/creepjs/')  # test fingerprint

input("Press ENTER to exit: ")
driver.quit()  # Execute on the End!

Don't forget to execute driver.quit() in the End. Else-wise your temporary folder will get flooded!

Run with Google-Colab

Google-Colab (file: master@google-colab/selenium_profiles.ipynb)

Profiles

Example Profile:

profile = \
{
  "options": {
      "sandbox": True,
      "headless": False,
      "load_images": True,
      "incognito": True,
      "touch": True,
      "app": False,
      "gpu": False,
      "proxy": "http://example-proxy.com:9000", # note: auth not supported,
      "extension_paths": ["path/to/extension_1", ...], # directory, .crx or .zip
      "args": ["--my-arg1", ...],
      "capabilities": {"cap_1":"val_1", "cap_2":"val_2"},
      "experimental_options":{"option1":"value1", "option2":"value2"},
      "adb": False, # run on harware device over ADB
      "adb_package": "com.android.chrome",
      "use_running_app": True
  },
  "cdp": {
    "touch": True,
    "darkmode":None,
    "maxtouchpoints": 5,
    "cores":8,
    "cdp_args": [],
    "emulation": {"mobile":True,"width": 384, "height": 700, "deviceScaleFactor": 10,
        "screenOrientation": {"type": "portrait-primary", "angle": 0}},
    "patch_version": True, # to patch automatically, or use "111.0.5563.111"
    "useragent": {
                "platform": "Linux aarch64",
                "acceptLanguage":"en-US",
                "userAgent": "Mozilla/5.0 (Linux; Android 11; HD1913) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/105.0.0.0 Mobile Safari/537.36",
                "userAgentMetadata": {
                    "brands": [{"brand": "Google Chrome", "version": "105"}, {"brand": "Not)A;Brand", "version": "8"},
                               {"brand": "Chromium", "version": "105"}],
                    "fullVersionList": [{"brand": "Google Chrome", "version": "105.0.5195.136"},
                                        {"brand": "Not)A;Brand", "version": "8.0.0.0"},
                                        {"brand": "Chromium", "version": "105.0.5195.136"}],
                    "fullVersion": "105.0.5195.136",
                    "platform": "Android",
                    "platformVersion": "11.0.0",
                    "architecture": "",
                    "model": "HD1913",
                    "mobile": True,
                    "bitness": "",
                    "wow64": False}
    }
  },
"proxy":{
  "proxy":"socks5://user1:pass@example_jost.com:5001", 
  "bypass_list":["localhost"]
  }
}

selenium-driverless

warning:

• this package is experimental and might include bugs, please report them at bug-reports
• might only work with 64-bit Python installations
• profiles might make it detected
• requires pip install selenium-driverless>=1.3.4 first

from selenium_profiles.webdriver import Chrome
from selenium_profiles.profiles import profiles
from selenium_driverless.webdriver import ChromeOptions
from selenium_driverless.types.by import By

profile = profiles.Windows()  # or .Android
options = ChromeOptions()
# options.add_argument("--headless=new")
driver = Chrome(profile, options=options, driverless_options=True)

# get url
driver.get('https://nowsecure.nl#relax')  # test fingerprint

driver.quit()  # Execute on the End!

see documentation for usages

Modify-headers

using selenium-wire

from selenium_profiles import webdriver
from selenium_profiles.profiles import profiles

profile = profiles.Android()

driver = webdriver.Chrome(profile, uc_driver=False, seleniumwire_options=True) # or pass seleniumwire-options

def interceptor(request):
    request.headers['New-Header'] = 'Some Value'
driver.request_interceptor = interceptor

# checkout headers
driver.get("https://httpbin.org/headers")

input("Press ENTER to quit..")
driver.quit()
exit()

Using Selenium-Injector

from selenium_profiles.webdriver import Chrome

driver = Chrome(injector_options=True)
injector = driver.profiles.injector

# modify headers
injector.declarativeNetRequest.update_headers({"test": "test_2", "sec-ch-ua-platform": "Android"})
rules = injector.declarativeNetRequest.dynamic_rules
headers = injector.declarativeNetRequest._headers

driver.get("https://httpbin.org/headers")
input("press ENTER to continue")

# block images
injector.declarativeNetRequest.update_block_on(resource_types=["image"])

driver.get("https://www.wikimedia.org/")

input("press ENTER to exit")
driver.quit()

Touch_actions

Example demonstration script

from selenium_profiles.webdriver import Chrome
from selenium_profiles.profiles import profiles
from selenium.webdriver.common.by import By
from selenium.webdriver import ChromeOptions

from selenium_profiles.scripts.driver_utils import TouchActionChain


# Start Driver
options = ChromeOptions()
profile = profiles.Android() # or .Windows()

driver = Chrome(profile, uc_driver=False, options=options)

# initialise touch_actions
chain = TouchActionChain(driver)

driver.get("https://cps-check.com/de/multi-touch-test")

touch_box = driver.find_element(By.XPATH,'//*[@id="box"]') # Get element



chain.touch_and_hold(touch_box)
chain.pause(10)
chain.release(touch_box)

# perform actions
chain.perform()

# now you should see a touch indication
# point on the Website for 10 seconds

# quit driver
input('Press ENTER to quit Driver\n')
driver.quit()

connect to running driver

Undetectability isn't garanteed

from selenium import  webdriver
driver = webdriver.Chrome()
# driver allready started:)


from selenium_profiles.webdriver import profiles as profile_manager
from selenium_profiles.profiles import profiles

profile = profiles.Android() # or .Android()
driver.profiles = profile_manager(driver=driver, profile=profile)
driver.profiles.apply(profile)

driver.get('https://hmaker.github.io/selenium-detector/')  # test fingerprint

input("Press ENTER to exit")
driver.quit()  # Execute on the End!

Set proxies dynamically or with options

from selenium_profiles.webdriver import Chrome
from selenium_profiles.profiles import profiles

profile = profiles.Windows() # or .Android()
profile["proxy"] = {
  "proxy":"http://user1:pass1@example_host.com:41149"
  }

driver = Chrome(profile=profile, injector_options=True)

driver.profiles.proxy.set_single("http://user2:pass2@example_host.com:41149")
print(driver.profiles.proxy.proxy)

driver.quit()  # Execute on the End!

To export a profile:

go to https://js.do/kaliiiiiiiiiii/get_profile in your browser and copy the text.

Help

Please feel free to open an issue or fork!

Known Bugs

• click_as_touch makes automation hung

Todo

• js-undetectability
  • [navigator.connection]
  • fonts don't match platform
  • does not match worker scope (Emulation) crbug#1358491
    • Navigator.userAgent
    • Navigator.platform
    • navigator.hardwareConcurrency
  • emulation leak on new tabs diskussion
  • selenium-detector
    • Either Devtools Console is open or CDP Runtime Domain is enabled => patch javascript objects using a Proxy or disable CDP.Runtime domain?
    • document.$cdc_asdjflasutopfhvcZLmcfl_
    • document.$chrome_asyncScriptInfo
    • driver.execute_script() usage (needs hook on called element)
    • driver.execute_async_script() usage (needs hook on called element)
    • driver.find_element() usage
    • window.cdc_adoQpoasnfa76pfcZLmcfl
• default metrics
  • Android
  • Windows
  • IOS
  • Linux
  • Tablet
• test.py script
  • test_driver.py
    • assert useragent, profile_export (no error)
      • Windows
        • useragent-data
        • undetected
          • headless
      • Android
      • useragent-data
      • undetected
        • headless

Deprecated

• Stealth method (Detected by Google)
• buster captcha solver | wontfix

Authors

Aurin Aegerter

License

Shield:

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Disclaimer

I am not responsible what you use the code for!!! Also no warranty!

Acknowledgments

Inspiration, code snippets, etc.

• Selenium
• selenium-documentation
• README-Template
• headless_js
• Selenium-Stealth
• Undetected-Chromedriver
• Selenium-Wire
• Modheader-Selenium
• ModHeader docs
• buster captcha solver | wontfix
• audio_captcha_solver
• Chromedriver-Options List
• Chrome DevTools Protocol (cdp_cmd)
• example_pypi_package
• google-colab installer
• scripts/touch_action_chain
• cdp_event_listeners
• proxy-auth
• webdriver-manager
• dynamic subclasses