trigger-argo-workflow resubmits the workflow multiple times after the workflow failure when --wait flag is used

[fornfrey]

When submitting a workflow with --wait flag, the process blocks until the workflow is completed. However, when the submitted workflow fails the action treats it as a transient error and attempts to resubmit it. This causes multiple AlreadyExists errors to be printed to the job output.

This does not impact the workflow submission or execution (the workflow failed anyway) but makes the debugging process confusing. Sample output of a failed job:

Notice: running command
  time: "2024-03-21T09:49:12.896Z"
  level: INFO
  extraArgs:
  - --wait
Error: failed to run command, retrying
  time: "2024-03-21T10:00:05.717Z"
  level: ERROR
  error: exit status 1
  retry_in: 559.753636ms
time="2024-03-21T10:00:06.993Z" level=fatal msg="Failed to submit workflow: rpc error: code = AlreadyExists desc = workflows.argoproj.io \"my-workflow-1\" already exists, the server was not able to generate a unique name for the object"
Error: failed to run command, retrying
  time: "2024-03-21T10:00:06.995Z"
  level: ERROR
  error: exit status 1
  retry_in: 1.077149047s
time="2024-03-21T10:00:08.773Z" level=fatal msg="Failed to submit workflow: rpc error: code = AlreadyExists desc = workflows.argoproj.io \"my-workflow-1\" already exists, the server was not able to generate a unique name for the object"