How does Meltano communicate with DBT after EL is complete How does Meltano instruct DBT to work only on a certain delta dataset instead of full table ?

If the end-user sets up a flow such as:

meltano run tap-jira target-snowflake dbt-snowflake:transform-jira

In this case

transform-jira

is a command set up like this: https://docs.meltano.com/concepts/project#plugin-commands You can set up as many commands as you want on top of the predefined ones. So if you want dbt to work only on new data, look to dbt to figure out how to do it through the existing CLI interface, then preconfigure that command in Meltano Meltano does not "instruct" dbt with such specificity as much as it is just running commands in subprocesses with as much specificity as the end-user encodes within them. Reduce or extend the above to your specific question