fix: batch transform failure handling

[CalebMartinUiPath]

failed batch transforms were being incorrectly handled as in progress. Causing the agent to loop waiting for it to finish. With this change agent execution will end properly.

[JeremyReist2]

does the same logic apply to DeepRAG?

[CalebMartinUiPath]

deeprag already has similar logic

[gcuip]

isn't the status redundant?

[gcuip]

why is this a user error? can't it be marked as failed due to other conditions in ecs?

in ecs we have:

try {
// do various async things
} catch (Exception ex)
            {
                ex = ex.Unwrap();

                if (info != null)
                {
                    var errorMessage = (ex as LocalizedException)?.Message
                                       ?? ErrorMessages.BatchRag_InternalFailure;
                    await Workflow.ExecuteLocalActivityAsync(
                        (IBatchRagReportingActivities a) => a.ReportFailureAsync(batchRagId, errorMessage),
                        new LocalActivityOptions
                        {
                            StartToCloseTimeout = TimeSpan.FromMinutes(1),
                            RetryPolicy = new RetryPolicy { MaximumAttempts = 3 }
                        });
                }
            }