Add inline_context check in LangChain

src/python/PythonSDK/foundationallm/langchain/agents/langchain_knowledge_management_agent.py

@@ -111,6 +111,7 @@ def _get_document_retriever(
         Get the vector document retriever, if it exists.
         """
         retriever = None
+
         if agent.vectorization is not None and not agent.inline_context:
             text_embedding_profile = AzureOpenAIEmbeddingProfile.from_object(
                 request.objects[agent.vectorization.text_embedding_profile_object_id]
@@ -241,7 +242,8 @@ def _validate_request(self, request: KnowledgeManagementCompletionRequest):
         self.prompt = self._get_prompt_from_object_id(request.agent.prompt_object_id, request.objects)
         if self.prompt.prefix is None or self.prompt.prefix == '':
             raise LangChainException("The Prompt object provided in the request's objects dictionary is invalid because it is missing a prefix value.", 400)
-        if request.agent.vectorization is not None:
+
+        if request.agent.vectorization is not None and not request.agent.inline_context:
             if request.agent.vectorization.text_embedding_profile_object_id is None or request.agent.vectorization.text_embedding_profile_object_id == '':
                 raise LangChainException("The TextEmbeddingProfileObjectId property of the agent's Vectorization property cannot be null or empty.", 400)