e2e-validate

End-to-end validation is needed when changes affect user workflows, UI, routing, authentication, or integrations, because unit tests miss regressions in these areas. It provides concrete, human-verifiable evidence that complete user flows still function correctly after modifications.

Run the project's end-to-end test runner to execute user workflow validation, then collect the generated playwright-report and test-results directories into .agent/artifacts/. This produces a PASS/FAIL result and an evidence string to attach to changelogs or feature trackers.

End-to-end validation operates without explicit dependencies, invoking the project's existing test runner. It requires a configured E2E environment like Playwright to execute workflows and generate the test outputs needed for artifact collection.

The best way to retain E2E test outputs is capturing test artifacts into .agent/artifacts/ or project logs. This records artifact pointers and an evidence string alongside PASS/FAIL results, ensuring human-verifiable evidence is retained for CI stage verification.

End-to-end validation should not be used for changes that do not impact user-facing behaviors or complete user workflows. It is designed for user flows, UI, routing, authentication, and integrations where human-verifiable evidence and retained artifacts are explicitly required.