router

Failed after

MSManuel Schiller
View latest

perf: skip unused Start request setup

8451001baa4Open on GitHub

Self-Healing CI

Classification: Flaky task

We classified this as a flaky test rather than a code change: the task has a 9% historical flakiness rate, the E2E project is not in the set of projects touched by this PR, and no similar failure was found in the comparison branch. The failure — a missing CSS computed style after Chromium client-side navigation — is consistent with a pre-existing non-deterministic rendering timing issue rather than a deterministic regression from these changes.

An empty commit was automatically applied to the branch to trigger a new CI pipeline execution to resolve the flaky task.

The failing test tanstack-solid-start-e2e-start-manifest:test:e2e was classified as a flaky task rather than a code change for the following reasons:

  1. Flakiness rate is non-zero (9%): Historical data shows this task fails intermittently ~9% of the time with the same code. This is a well-established signal of a pre-existing non-deterministic failure.

  2. The failing project is not in touched_projects: The E2E project tanstack-solid-start-e2e-start-manifest was not directly modified by this PR. The touched projects are root, @tanstack/start-client-core, @tanstack/start-server-core, and @tanstack/start-storage-context — none of which map to this E2E test suite.

  3. No similar failure in branch 8452: The similar-task-failure-detector found no matching failure in the comparison branch, ruling out a shared environment-state issue and indicating this is an isolated, intermittent occurrence.

  4. Browser CSS timing sensitivity: The test checks computed CSS styles (borderTopColor) after client-side navigation in Chromium. This class of assertion is particularly sensitive to rendering and hydration timing, making it prone to non-deterministic failures regardless of code changes.

  5. Direction-specific failure pattern is consistent with flakiness: The test fails only when navigating from lazy→static but passes for static→lazy. This asymmetric, one-off failure pattern is more consistent with a timing-related race in CSS hydration than a deterministic regression introduced by the PR.