Hi Scott,
A good way could be:
1. Define the "loop job" as a scheduled job, then turn off the schedule.
2. Define the "launcher job" in this way: on the last step, you can add a script step enabling the schedule of the "loop job" via
API (e.g: using curl) in an inline script.
Check
this curl example of that API endpoint.
In that way, the "loop job" should start the indefinite executions only after the execution of the "launcher job".
Hope it helps!