Context-Aware Networking & Runtimes: Agentic End-To-End

Source: Dev.to

AI network traffic can feel like a black box. You open an AI provider console or an Agent, ask a question or perform a task, and then what happens? Where does that traffic go? Is the traffic secure? Is it going to the appropriate destination? Where’s the context?

There are hundreds of questions that need to be answered from the moment an Agent makes a call to an LLM until you receive a response. To answer those questions you need an end‑to‑end workflow for when traffic leaves the Agent to when a task is completed.

Install Kagent

Kagent is an Agent framework that runs on Kubernetes.

Install the Kagent CRDs and create the kagent namespace

helm install kagent-crds oci://ghcr.io/kagent-dev/kagent/helm/kagent-crds \
    --namespace kagent \
    --create-namespace

Set your AI provider API key

For this example Anthropic is used, but any provider listed in the supported providers documentation will work.

export ANTHROPIC_API_KEY=your_api_key

Install Kagent with the chosen provider

helm upgrade --install kagent oci://ghcr.io/kagent-dev/kagent/helm/kagent \
    --namespace kagent \
    --set providers.default=anthropic \
    --set providers.anthropic.apiKey=$ANTHROPIC_API_KEY \
    --set ui.service.type=LoadBalancer

Verify the installation

kubectl get svc -n kagent

You should see output similar to:

Install Agentgateway + Kgateway

Kgateway is the control plane (the “brain”), and Agentgateway is the AI‑enabled data plane/proxy that handles all agentic traffic.

Install the Kubernetes Gateway API CRDs

kubectl apply -f https://github.com/kubernetes-sigs/gateway-api/releases/download/v1.4.0/standard-install.yaml

Install the Kgateway CRDs

helm upgrade -i --create-namespace --namespace kgateway-system kgateway-crds \
    oci://cr.kgateway.dev/kgateway-dev/charts/kgateway-crds \
    --version v2.1.1 \
    --set controller.image.pullPolicy=Always

Install Kgateway with Agentgateway enabled

helm upgrade -i -n kgateway-system kgateway \
    oci://cr.kgateway.dev/kgateway-dev/charts/kgateway \
    --version v2.1.1 \
    --set agentgateway.enabled=true \
    --set controller.image.pullPolicy=Always

Verify the control‑plane installation

kubectl get pods -n kgateway-system

List the available GatewayClasses

kubectl get gatewayclass

Expected output

NAME           CONTROLLER                  ACCEPTED   AGE
agentgateway   kgateway.dev/agentgateway   True       20h
kgateway       kgateway.dev/kgateway       True       20h

LLM Gateway Creation

With both the AI Agent framework (kagent) and the AI gateway (agentgateway) installed, you can now configure LLM‑related traffic.

Create a secret for your AI provider (Anthropic example)

apiVersion: v1
kind: Secret
metadata:
  name: anthropic-secret
  namespace: kagent
  labels:
    app: agentgateway
type: Opaque
stringData:
  Authorization: $ANTHROPIC_API_KEY

Apply the manifest:

kubectl apply -f -

Note: On a managed Kubernetes cluster the Gateway will receive a public IP. On a local cluster you may need to kubectl port-forward the service.

Define the Backend that points to the LLM

apiVersion: gateway.kgateway.dev/v1alpha1
kind: Backend
metadata:
  name: anthropic
  namespace: kgateway-system
  labels:
    app: agentgateway
spec:
  type: AI
  ai:
    llm:
      anthropic:
        authToken:
          kind: SecretRef
          secretRef:
            name: anthropic-secret
        model: "claude-3-5-haiku-latest"

Apply the manifest:

kubectl apply -f -

The :8080/anthropic endpoint will be routed securely to the Anthropic LLM, allowing you to observe and control AI‑related traffic end‑to‑end.