Don't flush the buffer and the sensitive data if we're reusing the output pipeline

Since we can reuse a single pipeline for multiple request-reply round
trips, we shouldn't flush the buffer and destroy the context.

Author: jhrozek

src/codegate/pipeline/output.py

@@ -160,6 +160,10 @@ async def process_stream(
             logger.error(f"Error processing stream: {e}")
             raise e
         finally:
+            # Don't flush the buffer if we assume we'll call the pipeline again
+            if cleanup_sensitive is False:
+                return
+
             # Process any remaining content in buffer when stream ends
             if self._context.buffer:
                 final_content = "".join(self._context.buffer)