chore(cleanup)!: remove tool_runtime.rag_tool

Kill the `builtin::rag` tool group completely since it is no longer
targeted. We use the Responses implementation for knowledge_search which
uses the `openai_vector_stores` pathway.

llama_stack/distributions/nvidia/run-with-safety.yaml

@@ -80,9 +80,7 @@ providers:
   scoring:
   - provider_id: basic
     provider_type: inline::basic
-  tool_runtime:
-  - provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+  tool_runtime: []
   files:
   - provider_id: meta-reference-files
     provider_type: inline::localfs
@@ -128,9 +126,7 @@ registered_resources:
   datasets: []
   scoring_fns: []
   benchmarks: []
-  tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  tool_groups: []
 server:
   port: 8321
 telemetry: