Question - We are running tekton, and it has "step pods" which essentially are running a shell script, each of the lines of output from that script are being collected by fluent in JSON as separate lines. Is there a way to use the multiline parser and collect all the output of each container without having to mark a start/end and have to update each script?