Greasy Fork镜像

changed many things, will update this if works.

No Redirects: We are not changing the page URL, so the server's check that caused the loop is irrelevant. We are winning the battle on their designated territory.
State Manipulation: Code is surgically altering the page, doing exactly what the site's own script would do after lets say 300 seconds.
Finds Hidden Elements: The key is the selector a.btn-primary[style*="display: none"]. Code is specifically looking for a button that the page has hidden from you.
Robust and Efficient: The MutationObserver ensures code runs at the exact moment the elements needed are available, making it fast and light on resources.

The script now only activates on the specific waiting page we want to skip.
@run-at document-start: This is crucial. It tells the browser to run our script the moment the HTML document begins to load, before the page's own scripts (like the timer) can even start.
No Observers, No Intervals: We don't need to wait for anything to appear. We are proactively taking control and redirecting the browser immediately.

Version 2.0 - This new script is event-driven. It does nothing until the page actually changes.

1. Because the MutationObserver is notified by the browser the instant the DOM changes, it will find and click the button immediately upon its appearance. setInterval could have a delay of up to a second after the button appears.

2. I've updated the @match rules to be more specific to the pages where downloads actually happen. This prevents the script from running unnecessarily on other pages of the site.
The DOWNLOAD_BUTTON_SELECTORS array makes the script more robust. If the site changes one button's class, you can easily add the new selector to the list without rewriting the logic. I've pre-filled it with the most likely candidates.

3. Once the button is found and clicked, observer.disconnect() is called. This cleans up after our script, so it doesn't continue to watch for changes forever, saving system resources.