We looked at the failing WebAuthn e2e test and think it is a timing flake, not something caused by our change. The test timed out waiting for the Logout button, and this PR only bumps the @microsoft/api-extractor build tool, which has no effect on journey-app runtime behaviour. We suggest rerunning the task.
Failed after Rrenovate[bot]View latest
chore(deps): update dependency @microsoft/api-extractor to v7.59.4
855de7643eOpen on GitHubSelf-Healing CI
Classification: Flaky task
An empty commit was automatically applied to the branch to trigger a new CI pipeline execution to resolve the flaky task.
The AI model provider was experiencing issues when this fix was generated.
Why this was not classified as a code_change:
- The error does not point at changed code.
- The failure is a Playwright assertion at
webauthn-device.test.ts:100. getByRole('button', { name: 'Logout' })was not visible within the 5000ms timeout after the authenticate step.- The message mentions no function, import or module touched by the PR.
- The PR is unrelated to the failing area.
- The only change is bumping the
@microsoft/api-extractordev dependency from 7.58.2 to 7.59.4. - This tool generates API reports and rolls up type declarations at build time.
- It does not affect journey-app runtime or WebAuthn behaviour.
- The PR diff lists no changed source files.
touched_projectscontains*only because of the lockfile and dependency change.
- The failure looks timing-related.
- The other test in the same file passed.
- The failing step is a browser UI flow that depends on a virtual authenticator and a render or redirect finishing within 5 seconds.
- That kind of step can pass or fail on timing alone.
- The task has a small flake rate.
- Its flakiness rate is about 1.4%, which is above 0%.
- That is consistent with an occasional non-deterministic failure.
- Removing the dependency bump would not logically fix the failure.
- Other checks.
- The task output was present, so the missing-output bail-out did not apply.
- No similar-failure branch was provided.
- No external-service outage is evident, so this is not environment_state.
Conclusion: this is a flaky_task. A rerun is the appropriate action, and no code changes are proposed.