Selenium: InvalidSelectorException: invalid selector

The driver could not use your locator because the selector itself is malformed or produces something that is not an element. It is a syntax problem, not a timing problem, so waiting longer will never fix it. Browsers report two main variants. For CSS: invalid selector: An invalid or illegal selector was specified. For XPath, a message that quotes your expression and says it is not a valid XPath expression, or one saying the result is [object Text] or [object Attr] and should be an element.

The XPath result variant is common in scraping code. Expressions ending in /text() or /@href are valid XPath, but WebDriver locators must return elements. Select the element and read .text or get_attribute('href') in code instead.

What is not wrong: the page. The same selector fails on every page.

from selenium.webdriver.common.by import By

# Invalid as a locator: returns attribute nodes
# driver.find_elements(By.XPATH, "//a[@class='product']/@href")

links = driver.find_elements(By.XPATH, "//a[@class='product']")
hrefs = [a.get_attribute("href") for a in links]

# Compound classes: use CSS, not CLASS_NAME
button = driver.find_element(By.CSS_SELECTOR, ".btn.btn-primary")

Validate selectors in DevTools before putting them in code, and keep locators in one place (page objects) so they are easy to test. Prefer short CSS selectors on stable attributes over long XPath expressions; reserve XPath for text matching and axis navigation that CSS cannot do.