gcloud-lab/apps/base/customer1/hermes-agent/deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: hermes-agent
  namespace: customer1
spec:
  replicas: 1
  selector:
    matchLabels:
      app: hermes-agent
  template:
    metadata:
      labels:
        app: hermes-agent
    spec:
      automountServiceAccountToken: false
      securityContext:
        runAsNonRoot: true
        runAsUser: 1000
        runAsGroup: 1000
        fsGroup: 1000
        seccompProfile:
          type: RuntimeDefault
      initContainers:
      - name: fix-webui-perms
        image: busybox:1.37
        imagePullPolicy: IfNotPresent
        securityContext:
          runAsNonRoot: true
          runAsUser: 1000
          runAsGroup: 1000
          allowPrivilegeEscalation: false
          capabilities:
            drop:
            - ALL
          seccompProfile:
            type: RuntimeDefault
        command:
          - sh
          - -c
        args:
          - |
            mkdir -p /data/.hermes/webui /data/.cache /data/.config /data/bin
            chmod -R g+rwX,o-rwx /data
            echo "✅ Hermes data permissions fixed (non-root with fsGroup)"
        volumeMounts:
          - name: hermes-data
            mountPath: /data
      containers:
      - name: hermes-agent
        securityContext:
          runAsNonRoot: true
          runAsUser: 1000
          runAsGroup: 1000
          allowPrivilegeEscalation: false
          readOnlyRootFilesystem: false
          capabilities:
            drop:
            - ALL
          seccompProfile:
            type: RuntimeDefault
        image: nousresearch/hermes-agent:latest
        command: ["/bin/bash", "-c"]
        args:
          - |
            set -euo pipefail
            echo "Hermes Agent starting (Telegram polling + full config persistence)..."

            HERMES_BIN="/opt/hermes/.venv/bin/hermes"

            if [ ! -x "$HERMES_BIN" ]; then
              echo "❌ hermes binary NOT FOUND at $HERMES_BIN"
              echo "Searching filesystem..."
              find / -name hermes -type f 2>/dev/null | head -10 || true
              exit 1
            fi

            echo "✅ Found hermes at: $HERMES_BIN"

            # Copy config
            cp -f /tmp/hermes/config.yaml /opt/data/config.yaml || true

            # Create symlink in a writable location (PVC) so it works when you exec in
            mkdir -p /opt/data/bin
            ln -sf "$HERMES_BIN" /opt/data/bin/hermes || true
            export PATH="/opt/data/bin:$PATH"

            echo "🚀 Starting hermes gateway..."
            exec "$HERMES_BIN" gateway run
        env:
          # === Hermes Home - REQUIRED for config persistence ===
          - name: HERMES_HOME
            value: "/opt/data"
          - name: HOME
            value: "/opt/data"
          - name: XDG_CACHE_HOME
            value: "/opt/data/.cache"
          - name: XDG_CONFIG_HOME
            value: "/opt/data/.config"

          # === Telegram Configuration (polling only - no public exposure) ===
          - name: TELEGRAM_BOT_TOKEN
            valueFrom:
              secretKeyRef:
                name: hermes-secrets
                key: TELEGRAM_BOT_TOKEN 
          - name: XAI_API_KEY
            valueFrom:
              secretKeyRef:
                name: xai-apikey
                key: XAI_API_KEY 
                
          - name: TELEGRAM_ALLOWED_USERS
            value: "7528130947"        

          # === Local vLLM (OpenAI-compatible) ===
          - name: OPENAI_BASE_URL
            value: "http://openclaw-brain-service.customer1.svc.cluster.local:8000/v1"   # ← adjust if your service name differs
          
          - name: HERMES_MODEL_PROVIDER
            value: xai
          
          - name: HERMES_MODEL 
            value: grok-4.20-0309-reasoning
          
          - name: OPENAI_API_KEY
            value: "dummy"                                 # vLLM ignores this

          # === API Server (for external OpenAI-compatible clients) ===
          - name: API_SERVER_ENABLED
            value: "true"
          - name: API_SERVER_HOST
            value: "0.0.0.0"
          - name: API_SERVER_PORT
            value: "8642"
          - name: API_SERVER_KEY
            value: ""  # empty = no auth (tailnet-only, private)
          - name: API_SERVER_MODEL_NAME
            value: "hermes-agent"

          # === Optional ===
          # - name: LOG_LEVEL
          #   value: "INFO"
        volumeMounts:
        - name: hermes-data
          mountPath: /opt/data
        - name: hermes-configmap
          mountPath: /tmp/hermes/config.yaml
          subPath: config.yaml
          readOnly: true
        resources:
          requests:
            memory: 2Gi
            cpu: "1"
          limits:
            memory: 4Gi
            cpu: "2"
     
      - name: hermes-webui
        image: ghcr.io/nesquena/hermes-webui:latest
        ports:
        - containerPort: 8787
        env:
          - name: HERMES_HOME
            value: "/home/hermeswebui/.hermes"
          - name: HERMES_WEBUI_HOST
            value: "0.0.0.0"
          - name: HERMES_WEBUI_PORT
            value: "8787"
          - name: HERMES_WEBUI_STATE_DIR
            value: "/home/hermeswebui/.hermes/webui"
          - name: WANTED_UID
            value: "1000"
          - name: WANTED_GID
            value: "1000"
        volumeMounts:
        - name: hermes-data
          mountPath: /home/hermeswebui/.hermes

        resources:
          requests:
            memory: 256Mi
            cpu: "100m"
          limits:
            memory: 512Mi
            cpu: "500m"
        securityContext:
          runAsNonRoot: true
          runAsUser: 1000
          runAsGroup: 1000
          allowPrivilegeEscalation: false
          readOnlyRootFilesystem: false
          capabilities:
            drop:
            - ALL
          seccompProfile:
            type: RuntimeDefault
      volumes:
      - name: hermes-configmap
        configMap:
          name: hermes-config
      - name: hermes-data
        persistentVolumeClaim:
          claimName: hermes-agent-pvc
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: hermes-agent-pvc
  namespace: customer1
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 25Gi