fix: set model id for custom models in llamastack config

[ganeshmurthy]

The llama-stack subchart requires each enabled model to have an id field for registered_resources.models.model_id. When using a custom model not predefined in the subchart, this field was never set, causing a pydantic ValidationError on startup. Default the id to the model key name, with an optional LLM_ID/SAFETY_ID override.

[ganeshmurthy]

Testing

Tested on OpenShift cluster with custom remote model:

  # Deployed with LLM_ID override
  make install \
    NAMESPACE=ai-virtual-agent \
    LLM=remote-llm \
    LLM_ID=Qwen3.6-35B-A3B \
    LLM_URL=https://litellm-litemaas.apps.prod.rhoai.rh-aiservices-bu.com/v1 \
    LLM_API_TOKEN=sk-***

Verification:

1. Helm values show correct override:
   $ helm get values ai-virtual-agent -n ai-virtual-agent | grep -A 5 "remote-llm:"

  remote-llm:
    apiToken: sk-***
    enabled: true
    id: Qwen3.6-35B-A3B  # ✅ Override applied, not "remote-llm"
    url: https://litellm-litemaas.apps.prod.rhoai.rh-aiservices-bu.com/v1

2. LlamaStack pod started successfully:

  $ oc get pods -n ai-virtual-agent | grep llamastack
  llamastack-79d47fb4f5-drnzv   1/1   Running   0   8m45s

- No validation errors (pod would crash if model_id field was missing or invalid)
- No restarts (clean startup with custom model configuration)

Result: The PR correctly allows LLM_ID to override the model identifier while LLM remains the YAML-safe config key. SAFETY model support also confirmed working.