Update more examples to use universal LLMContext. Specifically, update examples we didn't update before because they weren't using ToolsSchema for their tool definitions, which is a requirement for using LLMContext.

NOTE: oops! Turns out some of these files had *already* been updated to use universal `LLMContext` even though they weren't yet using `ToolsSchema`. This commit should fix those examples.

examples/foundational/33-gemini-rag.py

@@ -55,6 +55,8 @@ from dotenv import load_dotenv
 from google import genai
 from loguru import logger
 
+from pipecat.adapters.schemas.function_schema import FunctionSchema
+from pipecat.adapters.schemas.tools_schema import ToolsSchema
 from pipecat.audio.turn.smart_turn.base_smart_turn import SmartTurnParams
 from pipecat.audio.turn.smart_turn.local_smart_turn_v3 import LocalSmartTurnAnalyzerV3
 from pipecat.audio.vad.silero import SileroVADAnalyzer
@@ -63,7 +65,8 @@ from pipecat.frames.frames import LLMRunFrame
 from pipecat.pipeline.pipeline import Pipeline
 from pipecat.pipeline.runner import PipelineRunner
 from pipecat.pipeline.task import PipelineParams, PipelineTask
-from pipecat.processors.aggregators.openai_llm_context import OpenAILLMContext
+from pipecat.processors.aggregators.llm_context import LLMContext
+from pipecat.processors.aggregators.llm_response_universal import LLMContextAggregatorPair
 from pipecat.runner.types import RunnerArguments
 from pipecat.runner.utils import create_transport
 from pipecat.services.cartesia.tts import CartesiaTTSService
@@ -121,11 +124,7 @@ async def query_knowledge_base(params: FunctionCallParams):
 
     # for our case, the first two messages are the instructions and the user message
     # so we remove them.
-    conversation_turns = params.context.messages[2:]
-    # convert to standard messages
-    messages = []
-    for turn in conversation_turns:
-        messages.extend(params.context.to_standard_messages(turn))
+    conversation_turns = params.context.get_messages()[2:]
 
     def _is_tool_call(turn):
         if turn.get("role", None) == "tool":
@@ -135,7 +134,7 @@ async def query_knowledge_base(params: FunctionCallParams):
         return False
 
     # filter out tool calls
-    messages = [turn for turn in messages if not _is_tool_call(turn)]
+    messages = [turn for turn in conversation_turns if not _is_tool_call(turn)]
     # use the last 3 turns as the conversation history/context
     messages = messages[-3:]
     messages_json = json.dumps(messages, ensure_ascii=False, indent=2)
@@ -199,25 +198,20 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
         api_key=os.getenv("GOOGLE_API_KEY"),
     )
     llm.register_function("query_knowledge_base", query_knowledge_base)
-    tools = [
-        {
-            "function_declarations": [
-                {
-                    "name": "query_knowledge_base",
-                    "description": "Query the knowledge base for the answer to the question.",
-                    "parameters": {
-                        "type": "object",
-                        "properties": {
-                            "question": {
-                                "type": "string",
-                                "description": "The question to query the knowledge base with.",
-                            },
-                        },
-                    },
-                },
-            ],
+
+    query_function = FunctionSchema(
+        name="query_knowledge_base",
+        description="Query the knowledge base for the answer to the question.",
+        properties={
+            "question": {
+                "type": "string",
+                "description": "The question to query the knowledge base with.",
+            },
         },
-    ]
+        required=["question"],
+    )
+    tools = ToolsSchema(standard_tools=[query_function])
+
     system_prompt = """\
 You are a helpful assistant who converses with a user and answers questions.
 
@@ -230,8 +224,8 @@ Your response will be turned into speech so use only simple words and punctuatio
         {"role": "user", "content": "Greet the user."},
     ]
 
-    context = OpenAILLMContext(messages, tools)
-    context_aggregator = llm.create_context_aggregator(context)
+    context = LLMContext(messages, tools)
+    context_aggregator = LLMContextAggregatorPair(context)
 
     pipeline = Pipeline(
         [