I think it really depends on what external resource we’re interacting with. For something like BigQuery (which obviously stores data) we have sources and destinations for it. We plan to do this for other db/warehouse sources and destinations. For files on cloud storage we have a similar position.
For APIs there’s quite a bit larger range of what is possible. Some APIs don’t allow ingesting almost any data. Some, like Salesforce, allow ingesting data with a bit of configurability. Other APIs may only accept very limited options for data ingestion.
So in the general case, I would say that we do not say “_where there is a Source, there is also a Target_”. However, _we w_ill definitely add destinations for the highest profile sources that can ingest data. Eventually, we’ll likely improve our CDK to make it easier to implement these types of destinations as well.