feat: enable native XTTS splitting by default for better AI compatibility

app.py

@@ -1052,9 +1052,13 @@ def synthesize_speech(
             else:
                 return None, "Error: No audio segment could be generated", preprocessed_text
         else:
-            use_native_splitting = text_splitting_method == "Native XTTS splitting"
-            if len(preprocessed_text) > 150 and not use_native_splitting:
-                print("Long text detected: enabling native XTTS splitting to avoid truncation")
+            # Always enable native XTTS splitting by default for better AI agent compatibility
+            use_native_splitting = True
+            if text_splitting_method == "No splitting":
+                use_native_splitting = False
+                print("Native XTTS splitting disabled by user request")
+            elif len(preprocessed_text) > 150:
+                print("Long text detected: native XTTS splitting is enabled")
                 use_native_splitting = True
             
             print(f"Generating with parameters: temperature={temperature}, do_sample={do_sample}, repetition_penalty={repetition_penalty}, length_penalty={length_penalty}, top_k={top_k}, top_p={top_p}, enable_text_splitting={use_native_splitting}")