The configuration guide https://access.redhat.com/documentation/en/red-hat-jboss-enterprise-application-platform/7.0/paged/configuration-guide/chapter-20-configuring-batch-applications, in section 20.1.1 Configure Batch Job Repositories declares this:

******************************
Add a JDBC Job Repository

You can add a job repository that stores batch job information in a database. You must specify the data-source to connect to the database.

/subsystem=batch-jberet/jdbc-job-repository=REPOSITORY_NAME:add(data-source=java:jboss/datasources/DATASOURCE)
******************************

As you can see it points the repository to the JNDI name of the datasource. When we performed this action the result was:
******************************
[domain@192.168.0.132:9990 /] /profile=default/subsystem=batch-jberet/jdbc-job-repository=myreposotory:add(data-source=java:jboss/datasources/ExampleDS)
{
"outcome" => "failed",
"failure-description" =>

{"domain-failure-description" => "WFLYCTL0369: Required capabilities are not available: org.wildfly.data-source.java:jboss/datasources/ExampleDS in context 'profile=default'; There are no known registration points which can provide this capability."}

,
"rolled-back" => true
}
[domain@192.168.0.132:9990 /]
******************************

We pointed it to the pool-name instead of the JNDI-name of the datasource and that fixed the problem. I guess somebody should update the Configuration Guide.