Debugging Playwright Failures in CI Is Still Painful — I Tried to Fix It
The problem nobody talks about
Playwright gives you everything you need to debug a failed test:
- traces
- screenshots
- videos
- logs
In theory, debugging should be easy.
In practice, it’s not.
What actually happens in CI
When a test fails in CI, the workflow usually looks like this:
- download the trace
- open screenshots
- watch the video
- scroll through logs
- try to reconstruct what happened
All the data is there.
It’s just… scattered.
And the more tests you run, the worse this gets.
The real bottleneck
It’s not writing tests.
It’s not even flaky tests.
It’s this:
👉 figuring out why a test failed takes too long
Especially when you’re dealing with:
- parallel runs
- multiple environments
- CI pipelines
What I wanted instead
I didn’t want more data.
I wanted:
👉 everything about a failed test in one place
So I built a small open-source Playwright reporter
It collects everything from a test run and puts it into a single report:
- traces
- screenshots
- videos
- logs
No downloading artifacts.
No jumping between tools.
Just one place to understand what happened.
What it looks like
How it fits into a workflow
- Run tests (locally or in CI)
- Reporter collects artifacts
- Open one report → see everything
That’s it.
Optional: cloud debugging
If you're running tests in CI, there’s also an option to upload runs to a cloud dashboard (Sentinel) so you can inspect failures without downloading artifacts.
But the reporter itself works fully on its own.
Why I’m sharing this
I kept running into this problem over and over again, and I’m curious if others are dealing with the same thing.
How are you debugging Playwright failures in CI today?
Top comments (0)