fix: remove dead kubectl delete step from promote-to-uat workflow

The CTO correctly identified that the delete step was dead code:
- gcloud/kubectl silently fail in the runner (no GKE credentials)
- Architecturally wrong for GitOps (Flux handles reconciliation)
- Unique Job names + ttlSecondsAfterFinished handle lifecycle

Co-Authored-By: Paperclip <noreply@paperclip.ing>

.github/workflows/promote-to-uat.yml

@@ -72,25 +72,6 @@ jobs:
 
           git -C /tmp/infra diff --stat
 
-      - name: Delete existing seed Job in UAT (immutable Job fix)
-        env:
-          TAG: ${{ inputs.image_tag }}
-          GH_TOKEN: ${{ steps.infra-token.outputs.token }}
-        run: |
-          cd /tmp/infra
-          SHORT_SHA="${TAG##*-}"
-          SEED_JOB_NAME="seed-test-data-${SHORT_SHA}"
-          
-          echo "Deleting existing seed Job: ${SEED_JOB_NAME} in groombook-uat namespace"
-          
-          gcloud container clusters get-credentials groombook-uat --zone us-central1 --project groombook-424212 2>/dev/null || \
-          kubectl config view --minify --raw 2>/dev/null || true
-          
-          kubectl delete job/${SEED_JOB_NAME} -n groombook-uat --ignore-not-found=true 2>/dev/null || \
-          echo "Direct kubectl delete skipped (GitOps-only). Flux will reconcile after PR merge."
-          
-          echo "Job deletion complete. Flux will reconcile the new manifest after PR merge."
-
       - name: Create PR on groombook/infra
         env:
           TAG: ${{ inputs.image_tag }}