Retry
Domain: Automation & Scheduling
Motivation
Retry is how a failed or interrupted piece of automated work gets another chance. A job kill or failure settles the record; the model reads it with job_output and can start the work again. dsh-headless simply exits 1 and the outer script re-invokes the process; the webhook runtime is fire-and-forget with no built-in retry.
Concrete Example
After a background job settles as failed, job_list shows its status and a fresh start (or a second dsh —profile headless run) is the retry; schedule retries only when a later task-management change, another wake, or a Host restart re-evaluates the pending task.
Analogy
Rolling the dice again after a miss: the same bet, placed from scratch.