[FEATURE] Implement a UI Custom Resource Definition

[edenreich]

Summary

The Inference Gateway has a UI, it would be cool to also be able to deploy the UI if needed.

The spec could look something like this:

apiVersion: core.inference-gateway.com/v1alpha1
kind: UI
metadata:
  name: inference-gateway
  namespace: inference-gateway
spec:
  image: ghcr.io/inference-gateway/ui:latest
  environment: development
  port: 3000
  gatewayURL: http://inference-gateway.inference-gateway.svc.cluster.local:8080
  storage:
    type: local
  auth:
    enabled: true
    provider: keycloak
    oidc:
      issuerURL:
      issuerID:
      issuerSecret:
        secretKeyRef:
          name:
          key:
    cookies:
      secure: true

This will deploy a fully fledged UI that supports tool calls and streaming.

The UI project is here: https://github.com/inference-gateway/ui