Bugzilla

Another option for tests that use those chrome pages would be to run them isolated in a specific test job with the `disable_all_security` preference set to true - this is not a preference that we want to set in general for WebDriver! But by doing so it would allow us to as well get the extension installed in a branded build because the [check for `Cu.isInAutomation`](https://searchfox.org/mozilla-central/rev/dbef1a2f75798fb0136b7428d959c8feb09ad5d1/js/xpconnect/src/xpcpublic.h#874-880) will return `true`.

Another option for tests that use those chrome pages would be to run them isolated in a specific test job with the `disable_all_security` preference set to true - this is not a preference that we want to set in general for WebDriver and is required to be set before Firefox starts! 

But when setting the preference, it would allow us to as well get the extension installed in a branded build because the [check for `Cu.isInAutomation`](https://searchfox.org/mozilla-central/rev/dbef1a2f75798fb0136b7428d959c8feb09ad5d1/js/xpconnect/src/xpcpublic.h#874-880) will then return `true`.

Another option for tests that use those chrome pages would be to run them isolated in a specific test job with the `disable_all_security` preference set to true - this is not a preference that we want to set in general for WebDriver and is required to be set before Firefox starts! 

But when setting the preference, it would allow us to as well get the extension installed in a branded build because the [check for `Cu.isInAutomation`](https://searchfox.org/mozilla-central/rev/dbef1a2f75798fb0136b7428d959c8feb09ad5d1/js/xpconnect/src/xpcpublic.h#874-880) will then return `true`.

However, this approach would bypass certain security checks, which risks masking failures or regressions that real Firefox users might encounter when automating with WebDriver. That said, I'm not necessarily opposed to taking that route.