Remove unnecessary system_instruction argument from run_inference() methods

src/pipecat/pipeline/llm_switcher.py

@@ -30,25 +30,17 @@ class LLMSwitcher(ServiceSwitcher[StrategyType]):
         """Get the currently active LLM, if any."""
         return self.strategy.active_service
 
-    async def run_inference(
-        self, context: LLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context, using the currently active LLM.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context. If both are provided, the
-              one in the context takes precedence.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.
         """
         if self.active_llm:
-            return await self.active_llm.run_inference(
-                context=context, system_instruction=system_instruction
-            )
+            return await self.active_llm.run_inference(context=context)
         return None
 
     def register_function(

src/pipecat/services/anthropic/llm.py

@@ -228,17 +228,11 @@ class AnthropicLLMService(LLMService):
             response = await api_call(**params)
             return response
 
-    async def run_inference(
-        self, context: LLMContext | OpenAILLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext | OpenAILLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context. If both are provided, the
-              one in the context takes precedence.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.
@@ -255,7 +249,7 @@ class AnthropicLLMService(LLMService):
         else:
             context = AnthropicLLMContext.upgrade_to_anthropic(context)
             messages = context.messages
-            system = getattr(context, "system", None) or system_instruction or NOT_GIVEN
+            system = getattr(context, "system", NOT_GIVEN)
 
         # LLM completion
         response = await self._client.messages.create(

src/pipecat/services/aws/llm.py

@@ -792,17 +792,11 @@ class AWSBedrockLLMService(LLMService):
         """
         return True
 
-    async def run_inference(
-        self, context: LLMContext | OpenAILLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext | OpenAILLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context. If both are provided, the
-              one in the context takes precedence.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.
@@ -822,7 +816,7 @@ class AWSBedrockLLMService(LLMService):
             else:
                 context = AWSBedrockLLMContext.upgrade_to_bedrock(context)
                 messages = context.messages
-                system = getattr(context, "system", None) or system_instruction
+                system = getattr(context, "system", None)
 
             # Determine if we're using Claude or Nova based on model ID
             model_id = self.model_name

src/pipecat/services/google/llm.py

@@ -733,17 +733,11 @@ class GoogleLLMService(LLMService):
     def _create_client(self, api_key: str, http_options: Optional[HttpOptions] = None):
         self._client = genai.Client(api_key=api_key, http_options=http_options)
 
-    async def run_inference(
-        self, context: LLMContext | OpenAILLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext | OpenAILLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context. If both are provided, the
-              one in the context takes precedence.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.
@@ -758,7 +752,7 @@ class GoogleLLMService(LLMService):
         else:
             context = GoogleLLMContext.upgrade_to_google(context)
             messages = context.messages
-            system = getattr(context, "system_message", None) or system_instruction
+            system = getattr(context, "system_message", None)
 
         generation_config = GenerateContentConfig(system_instruction=system)
 

src/pipecat/services/llm_service.py

@@ -195,18 +195,13 @@ class LLMService(AIService):
         """
         return self._adapter
 
-    async def run_inference(
-        self, context: LLMContext | OpenAILLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext | OpenAILLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context.
 
         Must be implemented by subclasses.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.

src/pipecat/services/openai/base_llm.py

@@ -245,16 +245,11 @@ class BaseOpenAILLMService(LLMService):
         params.update(self._settings["extra"])
         return params
 
-    async def run_inference(
-        self, context: LLMContext | OpenAILLMContext, system_instruction: Optional[str] = None
-    ) -> Optional[str]:
+    async def run_inference(self, context: LLMContext | OpenAILLMContext) -> Optional[str]:
         """Run a one-shot, out-of-band (i.e. out-of-pipeline) inference with the given LLM context.
 
         Args:
             context: The LLM context containing conversation history.
-            system_instruction: Optional system instruction to guide the LLM's
-              behavior. You could also (again, optionally) provide a system
-              instruction directly in the context.
 
         Returns:
             The LLM's response as a string, or None if no response is generated.