[Bug]: e2e tests have become flaky

[MatissJanis]

Verified issue does not already exist?

• I have searched and found no existing issue
• I will be providing steps how to reproduce the bug (in most cases this will also mean uploading a demo budget file)

What happened?

The e2e tests have become increasingly more flaky in the past few days. Need to look into what's the issue and address it.

Where are you hosting Actual?

None

What browsers are you seeing the problem on?

No response

Operating System

None

[joel-jeremy]

One this I noticed is that the rules page is crashing on VRT but when you open the page in the browser it works just fine.
VRT:

Browser:

[joel-jeremy]

This change also appears randomly:

[UnderKoen]

One thing I noticed with #3499 was that the css classes of glamour weren't applied with the same precedence as in the browser. This can be caused by the order of the css in the final css changing

[MatissJanis]

Another issue I'm seeing semi-frequently: onboarding test.

---

Sample failing ci job: https://github.com/actualbudget/actual/actions/runs/11037117182/job/30657293317

Trace: trace.zip

View the trace here: https://trace.playwright.dev

---

Edit: it seems that there are a bunch of different redirects happening when clicking "start fresh". Some are redirecting to /budget and some to /accounts. Must be a race condition somewhere.

[MatissJanis]

Both the rules & onboarding e2e should be more stable now. Haven't been able to reproduce the settings one yet. If anyone sees a failing CI job with it - please post it here.

Or if you also see other e2e ci jobs failing (that use the latest changes in master) - please post them here.

[MatissJanis]

Related #3506