support testing malicious prompts with no guidelines

tests/security/garak.rest.llm-rag.json

@@ -2,7 +2,7 @@
     "rest": {
         "RestGenerator": {
             "name": "Microsoft ONNX Phi-3 LLM/GenAI REST API Service (RAG)",
-            "uri": "http://localhost:9999/api/rag_conversations",
+            "uri": "http://localhost:9999/api/completions/rag-guided",
             "method": "post",
             "headers": {
                 "Content-Type": "application/json"

tests/security/garak.rest.llm.json

@@ -2,7 +2,7 @@
     "rest": {
         "RestGenerator": {
             "name": "Microsoft ONNX Phi-3 LLM/GenAI REST API Service",
-            "uri": "http://localhost:9999/api/conversations",
+            "uri": "http://localhost:9999/api/completions",
             "method": "post",
             "headers": {
                 "Content-Type": "application/json"

tests/security/run_garak.sh

@@ -1,5 +1,5 @@
 garak -v \
     --config ./garak.config.yml \
-    --generator_option_file ./garak.rest.llm-rag.json \
+    --generator_option_file ./garak.rest.llm.json \
     --model_type=rest \
     --parallel_attempts 32