feat: Add LOCAL variable support for responses with UTF-8 JSON encoding.

Description: This feature adds support for UTF-8 encoding in JSON responses, allowing for proper handling of special characters and emojis.

- Implement unit tests for UTF-8 encoding in various model providers including Gemini, OpenAI, and OpenAI Compatible.
- Validate UTF-8 support in token counting, content generation, and error handling.
- Introduce tests for JSON serialization ensuring proper handling of French characters and emojis.
- Create tests for language instruction generation based on locale settings.
- Validate UTF-8 handling in workflow tools including AnalyzeTool, CodereviewTool, and DebugIssueTool.
- Ensure that all tests check for correct UTF-8 character preservation and proper JSON formatting.
- Add integration tests to verify the interaction between locale settings and model responses.

tools/consensus.py

@@ -512,7 +512,10 @@ of the evidence, even when it strongly points in one direction.""",
                     "provider_used": provider.get_provider_type().value,
                 }
 
-                return [TextContent(type="text", text=json.dumps(response_data, indent=2))]
+                return [TextContent(
+                    type="text", 
+                    text=json.dumps(response_data, indent=2, ensure_ascii=False)
+                )]
 
         # Otherwise, use standard workflow execution
         return await super().execute_workflow(arguments)

tools/shared/base_tool.py

@@ -1067,6 +1067,22 @@ Consider requesting searches for:
 
 When recommending searches, be specific about what information you need and why it would improve your analysis. Always remember to instruct Claude to use the continuation_id from this response when providing search results."""
 
+    def get_language_instruction(self) -> str:
+        """
+        Generate language instruction based on LOCALE configuration.
+
+        Returns:
+            str: Language instruction to prepend to prompt, or empty string if
+                 no locale set
+        """
+        from config import LOCALE
+
+        if not LOCALE or not LOCALE.strip():
+            return ""
+
+        # Simple language instruction
+        return f"Always respond in {LOCALE.strip()}.\n\n"
+
     # === ABSTRACT METHODS FOR SIMPLE TOOLS ===
 
     @abstractmethod

tools/simple/base.py

@@ -372,24 +372,24 @@ class SimpleTool(BaseTool):
 
                 follow_up_instructions = get_follow_up_instructions(0)
                 prompt = f"{prompt}\n\n{follow_up_instructions}"
-                logger.debug(f"Added follow-up instructions for new {self.get_name()} conversation")
-
-            # Validate images if any were provided
+                logger.debug(f"Added follow-up instructions for new {self.get_name()} conversation")            # Validate images if any were provided
             if images:
                 image_validation_error = self._validate_image_limits(
                     images, model_context=self._model_context, continuation_id=continuation_id
                 )
                 if image_validation_error:
-                    return [TextContent(type="text", text=json.dumps(image_validation_error))]
+                    return [TextContent(
+                        type="text", 
+                        text=json.dumps(image_validation_error, ensure_ascii=False)
+                    )]
 
             # Get and validate temperature against model constraints
             temperature, temp_warnings = self.get_validated_temperature(request, self._model_context)
 
             # Log any temperature corrections
             for warning in temp_warnings:
+                # Get thinking mode with defaults
                 logger.warning(warning)
-
-            # Get thinking mode with defaults
             thinking_mode = self.get_request_thinking_mode(request)
             if thinking_mode is None:
                 thinking_mode = self.get_default_thinking_mode()
@@ -398,7 +398,9 @@ class SimpleTool(BaseTool):
             provider = self._model_context.provider
 
             # Get system prompt for this tool
-            system_prompt = self.get_system_prompt()
+            base_system_prompt = self.get_system_prompt()
+            language_instruction = self.get_language_instruction()
+            system_prompt = language_instruction + base_system_prompt
 
             # Generate AI response using the provider
             logger.info(f"Sending request to {provider.get_provider_type().value} API for {self.get_name()}")

tools/workflow/workflow_mixin.py

@@ -715,7 +715,10 @@ class BaseWorkflowMixin(ABC):
             if continuation_id:
                 self.store_conversation_turn(continuation_id, response_data, request)
 
-            return [TextContent(type="text", text=json.dumps(response_data, indent=2))]
+            return [TextContent(
+                type="text", 
+                text=json.dumps(response_data, indent=2, ensure_ascii=False)
+            )]
 
         except Exception as e:
             logger.error(f"Error in {self.get_name()} work: {e}", exc_info=True)
@@ -728,7 +731,10 @@ class BaseWorkflowMixin(ABC):
             # Add metadata to error responses too
             self._add_workflow_metadata(error_data, arguments)
 
-            return [TextContent(type="text", text=json.dumps(error_data, indent=2))]
+            return [TextContent(
+                type="text", 
+                text=json.dumps(error_data, indent=2, ensure_ascii=False)
+            )]
 
     # Hook methods for tool customization
 
@@ -1233,7 +1239,7 @@ class BaseWorkflowMixin(ABC):
         # - file_context (internal optimization info)
         # - required_actions (internal workflow instructions)
 
-        return json.dumps(clean_data, indent=2)
+        return json.dumps(clean_data, indent=2, ensure_ascii=False)
 
     # Core workflow logic methods
 
@@ -1265,7 +1271,10 @@ class BaseWorkflowMixin(ABC):
                 # Promote the special status to the main response
                 special_status = expert_analysis["status"]
                 response_data["status"] = special_status
-                response_data["content"] = expert_analysis.get("raw_analysis", json.dumps(expert_analysis))
+                response_data["content"] = expert_analysis.get(
+                    "raw_analysis", 
+                    json.dumps(expert_analysis, ensure_ascii=False)
+                )
                 del response_data["expert_analysis"]
 
                 # Update next steps for special status
@@ -1524,20 +1533,22 @@ class BaseWorkflowMixin(ABC):
                 error_data = {"status": "error", "content": "No arguments provided"}
                 # Add basic metadata even for validation errors
                 error_data["metadata"] = {"tool_name": self.get_name()}
-                return [TextContent(type="text", text=json.dumps(error_data))]
+                return [TextContent(
+                    type="text", 
+                    text=json.dumps(error_data, ensure_ascii=False)
+                )]
 
             # Delegate to execute_workflow
             return await self.execute_workflow(arguments)
 
         except Exception as e:
             logger.error(f"Error in {self.get_name()} tool execution: {e}", exc_info=True)
-            error_data = {"status": "error", "content": f"Error in {self.get_name()}: {str(e)}"}
-            # Add metadata to error responses
+            error_data = {"status": "error", "content": f"Error in {self.get_name()}: {str(e)}"}            # Add metadata to error responses
             self._add_workflow_metadata(error_data, arguments)
             return [
                 TextContent(
                     type="text",
-                    text=json.dumps(error_data),
+                    text=json.dumps(error_data, ensure_ascii=False),
                 )
             ]