Advanced AI Prompt for Multi-Region Zero-Downtime Failover Pipelines

Act as a Global High-Availability Architect. Design a complex, multi-region failover GitHub Actions workflow for an active-active enterprise service. The pipeline must sequence the rollout to mitigate global outages. Stage 1: Deploy exclusively to [PRIMARY_REGION]. Stage 2: Execute deep integration tests against the primary region. Stage 3: Upon success, execute identical deployment to [SECONDARY_REGION]. Stage 4: The critical routing phase. The workflow must execute AWS CLI commands to manipulate Route 53 weighted records within [ROUTE53_HOSTED_ZONE], gradually shifting traffic from Primary to Secondary. The script must pause and poll the [HEALTH_CHECK_URL] from an external vantage point; if latency exceeds 500ms or 5xx errors occur, immediately revert the DNS weighting and fail the deployment build.