At the moment process looks like this:
I have a quarkus project and a serverless workflow json (see serverless.org for details) file inside. For each of different wofkflows I have to fork the project and/or rebuild it with another serverless worflow json which is put into resources.
Is there a way to have a "generic" kogito instance which can read serverless workflow json at runtime and run a workflow for me?
You can run serverless workflow definitions (either written in json/yaml or programatically generated using Java fluent API) embedded into a JVM. See https://sonataflow.org/serverlessworkflow/main/getting-started/java-embedded-workflows.html for more details on that.