devenv/tests/automation/load_test.go:87

• With the t.Error removal, scanErr won’t cause t.Failed() to be true until the final require.NoError, so the log-saving branch is skipped when log scanning fails. Save container logs when scanErr != nil (e.g., if t.Failed() || scanErr != nil) so failures in the log scanner still capture logs for debugging.

		if t.Failed() {
			saveErr := framework.StreamCTFContainerLogsFanout(
				framework.LogStreamConsumer{
					Name: "save-container-logs",
					Consume: func(logStreams map[string]io.ReadCloser) error {
						_, saveErr := framework.SaveContainerLogsFromStreams(fmt.Sprintf("%s-%d", framework.DefaultCTFLogsDir, time.Now().UnixNano()), logStreams)
						return saveErr

devenv/tests/ocr2/soak_test.go:52

• Removing the early t.Error means a non-nil scanErr no longer marks the test as failed until the final require.NoError, so the if t.Failed() branch won’t run and container logs won’t be saved for log-scan failures. Consider saving logs when scanErr != nil as well (e.g., if t.Failed() || scanErr != nil), and update the final error message to reflect a scan failure rather than a save failure.

		if t.Failed() {
			saveErr := framework.StreamCTFContainerLogsFanout(
				framework.LogStreamConsumer{
					Name: "save-container-logs",
					Consume: func(logStreams map[string]io.ReadCloser) error {