A multi-agent release workflow failed after Copilot cloud agent pushed a remediation branch and a validation agent reported degraded canary metrics. The team expected the recovery workflow to open an incident and wait for a human-triggered rollback. Which option best explains why the rollback did not execute? Select one answer.

The rollback-production job was skipped because its if condition evaluated to false. The condition github.eventname == 'workflowdispatch' && inputs.targetsha != '' indicates that this job is designed to run only when the workflow is manually triggered (workflowdispatch) and a specific targetsha input is provided. Since the recovery workflow was triggered by a workflowrun event (as shown by Event: workflowrun) and not a manual dispatch, the condition was not met, and the rollback did not execute automatically. The team's expectation for a human-triggered rollback aligns with this configuration, as it requires manual intervention to initiate the workflowdispatch event. The other options are incorrect because: The deployment secret being unavailable to forks is irrelevant as the job didn't even start to evaluate secrets. Required reviewer rejection is not applicable because deployment protection rules were not evaluated due to the job being skipped. The checks permission blocking incident creation is incorrect, as the create-incident-issue job successfully created issue 418.