Every pipeline has its own pipeline.json file located in the data directory. One could potentially write a script to update each pipeline.json file which has the Kafka stage lib. For example, you could replace all occurrences of streamsets-datacollector-apache-kafka_0_9-lib with streamsets-datacollector-apache-kafka_0_10-lib.
There is currently a feature request open to do this from StreamSets Control Hub, but we don’t have any plans for Data Collector at the moment.
