Failed after

SGSarah Gerrard
View latest

perf(router-plugin): compile code-splitting grouping validation

8301ddfc3f8Open on GitHub

Self-Healing CI

Classification: Environment state

We classified both CI failures as pre-existing environment issues unrelated to this PR's changes. The @tanstack/start-plugin-core:test:unit snapshot mismatches are identical to failures already present in the base branch (8300), and the tanstack-react-start-e2e-basic e2e failure is caused by a stale e2e-utils build artifact missing the collectBrowserErrors export. Neither failure is in a project touched by this PR.

An empty commit was applied to the branch to trigger a new CI pipeline execution to resolve the environment issue.

Both failing tasks are classified as environment_state because neither is caused by the PR's changes.

Task 1: @tanstack/start-plugin-core:test:unit

  • The failing project (@tanstack/start-plugin-core) is not in the PR's touched projects (root, @tanstack/router-plugin).
  • The PR only wraps an existing Zod schema with z.compile() in packages/router-plugin/src/core/config.ts — it does not touch any start-plugin-core source files.
  • The 8 snapshot mismatches in tests/hydrateWhen/hydrateWhen.test.ts show JSX indentation differences that are identical to failures already present in branch 8300 (the base branch this PR is stacked on), confirming these are pre-existing failures not introduced by this PR.

Task 2: tanstack-react-start-e2e-basic:test:e2e--rsbuild-ssr--iife

  • The failing project (tanstack-react-start-e2e-basic) is also not in the PR's touched projects.
  • The error is SyntaxError: The requested module './fixture.js' does not provide an export named 'collectBrowserErrors', which originates from e2e/e2e-utils/dist/esm/index.js — a compiled build artifact.
  • This indicates the e2e-utils package was not built (or was built from an older version) that does not expose collectBrowserErrors, a build/environment state problem entirely unrelated to the schema compilation change in this PR.
  • This failure does not appear in branch 8300, further confirming it is a local environment build state issue rather than a regression introduced by the PR's code changes.