Expose same endpoints provided by Confluent's Schema Registry. For more details on API endpoints, click here.
It's meant to allow a pluggable datastore, for whoever that needs a Schema Registry like functionality but don't want to run Kafka. For example, if you are data architecture is based on Kinesis and you would like to store schemas on a datastore provided by AWS like a relational db (RDS), elasticache (Redis), etc.
By the way, Schema Registry from Confluent is an amazing product which you should definetely consider if you are running Kafka.