Demo
Sign in

run health + AI insights render

tests/e2e/dashboard/run-health.spec.ts

failed
1m 29s
chromium
Diagnosis

See the failing step first, then validate the likely cause with the trace and artifacts.

Failure summary
AssertionError
High confidence
Expected metric pass rate to show 89% but the dashboard rendered 88%.
This is the kind of failure Sentinel should explain immediately: the UI is stable, the expected text is explicit, and the mismatch is visible in the screenshot.
Parsed failure signal
AssertionError: Expected metric pass rate to show 89% but the dashboard rendered 88%.
Evidence
  • Playwright assertion captured both expected and received text values.
  • The screenshot shows a completed dashboard state, not a loading or auth interruption.
  • The failing locator is specific enough to point straight at the broken assertion.
What to check next
  • Confirm whether the product contract changed from 89% to 88%.
  • Review the assertion line before assuming there is a timing issue.
  • Use trace only to confirm the metric never changed during the run.
Suggested fix
If 88% is the intended product state, update the assertion to the current UI contract.
await expect(page.getByTestId('metric-pass-rate')).toHaveText('88%');
Loading failing step…
Evidence

Open the same artifacts a developer would inspect in a real failing run.

Evidence workspace

Best first click for step timing and action flow.

Scripted demo
Trace timeline
21:43:08 Navigate /ops-room
21:43:10 Wait for dashboard cards
21:43:14 Assert pass-rate metric is 89%
21:43:19 Assertion reads visible value 88%
21:43:19 Test fails
Start with trace
Use the trace to confirm the metric stayed stable throughout the run.
Screenshot and logs
The screenshot should make the mismatch obvious at a glance.
Logs matter less here unless a backend recalculation should have happened.
Advanced analysis

This is where the product proves it is more than an artifact viewer.

Failure intelligence

Computing failure intelligence…