Neb Debug Quick

CRITICAL: AI models MUST NOT re-enable selfHeal or prune. Only humans restore these settings.

2. Enter debug mode

The enter command patches the workload to:

• Remove liveness, readiness, startup probes
• Set readOnlyRootFilesystem: false
• Set runAsUser: 0, runAsGroup: 0, runAsNonRoot: false (run as root for full filesystem write access)

scripts/neb-debug-quick.sh enter \
  --cluster infra-use2-mgmt-aws-0 \
  --ns infra-use2-mgmt-aws-0-platform-backend \
  --workload deploy/platform-backend

3. Build and sync

# Go services (cross-compile for arm64)
GOOS=linux GOARCH=arm64 go build -o ./bin/platform-backend ./cmd/platform/main.go
scripts/neb-debug-quick.sh sync \
  --cluster infra-use2-mgmt-aws-0 \
  --ns infra-use2-mgmt-aws-0-platform-backend \
  --selector app.kubernetes.io/name=platform-backend \
  --src ./bin/platform-backend --dst /platform-backend

4. Restart the service

scripts/neb-debug-quick.sh restart \
  --cluster infra-use2-mgmt-aws-0 \
  --ns infra-use2-mgmt-aws-0-platform-backend \
  --selector app.kubernetes.io/name=platform-backend

5. Restore

scripts/neb-debug-quick.sh restore \
  --cluster infra-use2-mgmt-aws-0 \
  --ns infra-use2-mgmt-aws-0-platform-backend \
  --workload deploy/platform-backend

Rules

• NEVER re-enable selfHeal or prune programmatically — only humans do this
• NEVER trigger CI builds — all code changes are built locally and pushed to pods via sync
• NEVER rebuild container images — use sync + restart instead
• NEVER restart/delete/recreate pods — use supervisorctl restart app via restart command
• NEVER use kubectl rollout restart, kubectl delete pod, or any pod lifecycle commands
• NEVER use local argocd CLI — always exec into the argocd-server pod
• NEVER use argocd-server API directly — use CLI inside the pod
• NEVER use kubectl port-forward for argocd
• NEVER call the AI server directly — always use platform-backend proxy

The debug cycle is: build locally → sync to pod → supervisorctl restart app. No CI, no image builds, no pod restarts.

Common Patterns

Debug platform-backend

cd nebcore/platform-backend
GOOS=linux GOARCH=arm64 go build -o ./bin/platform-backend ./cmd/platform/main.go
scripts/neb-debug-quick.sh sync --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-platform-backend --selector app.kubernetes.io/name=platform-backend --src ./bin/platform-backend --dst /platform-backend
scripts/neb-debug-quick.sh restart --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-platform-backend --selector app.kubernetes.io/name=platform-backend

Debug nebcore-ai-server (Node.js)

scripts/neb-debug-quick.sh sync --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-nebcore-ai-server --selector app.kubernetes.io/name=nebcore-ai-server --src ./patches/activity-log.js --dst /app/server/dist/services/activity-log.js
scripts/neb-debug-quick.sh restart --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-nebcore-ai-server --selector app.kubernetes.io/name=nebcore-ai-server

Debug nebcore-operator

cd nebcore/nebcore-operator
GOOS=linux GOARCH=arm64 go build -o ./bin/nebcore-operator .
scripts/neb-debug-quick.sh sync --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-nebcore-operator --selector app.kubernetes.io/name=nebcore-operator --src ./bin/nebcore-operator --dst /nebcore-operator
scripts/neb-debug-quick.sh restart --cluster infra-use2-mgmt-aws-0 --ns infra-use2-mgmt-aws-0-nebcore-operator --selector app.kubernetes.io/name=nebcore-operator