llamastack/llama-stack-k8s-operator

AI Summary: Create a new shell script, `hack/deploy-vllm.sh`, that deploys a vLLM inference server to Kubernetes. The script should allow configuration of the vLLM image, include a health check to ensure the deployment is ready, and accept a parameter for specifying the model to use. This will mirror the existing `hack/deploy-ollama.sh` script, providing a streamlined deployment experience for vLLM users.

AI Summary: The task involves refactoring the Llama Stack Kubernetes operator to use `cli.Path` instead of `cli.Update` for applying Kubernetes resources, improving code consistency and potentially performance. This is part of a larger effort to standardize resource application methods within the operator.