fix(scene): fix chat scene config custom prompt (#2725)

packages/dbgpt-app/src/dbgpt_app/scene/base_chat.py

@@ -172,7 +172,18 @@ class BaseChat(ABC):
             prompt_template = self._prompt_service.get_template(self.prompt_code)
             chat_prompt_template = ChatPromptTemplate(
                 messages=[
-                    SystemPromptTemplate.from_template(prompt_template.template),
+                    SystemPromptTemplate.from_template(
+                        template=prompt_template.template,
+                        template_format=prompt_template.template_format,
+                        response_format=(
+                            prompt_template.response_format
+                            if prompt_template.response_format
+                            and prompt_template.response_format != "{}"
+                            else None
+                        ),
+                        response_key=prompt_template.response_key,
+                        template_is_strict=prompt_template.template_is_strict,
+                    ),
                     MessagesPlaceholder(variable_name="chat_history"),
                     HumanPromptTemplate.from_template("{question}"),
                 ]
@@ -212,6 +223,30 @@ class BaseChat(ABC):
         """
         return self.parse_user_input()
 
+    async def prepare_input_values(self) -> Dict:
+        """Generate input value compatible with custom prompt to LLM
+
+        Please note that you must not perform any blocking operations in this function
+
+        Returns:
+            a dictionary to be formatted by prompt template
+        """
+        input_values = await self.generate_input_values()
+
+        # Mapping variable names: compatible with custom prompt template variable names
+        # Get the input_variables of the current prompt
+        input_variables = []
+        if hasattr(self.prompt_template, "prompt") and hasattr(
+            self.prompt_template.prompt, "input_variables"
+        ):
+            input_variables = self.prompt_template.prompt.input_variables
+        # Compatible with question and user_input
+        if "question" in input_variables and "question" not in input_values:
+            input_values["question"] = self.current_user_input
+        if "user_input" in input_variables and "user_input" not in input_values:
+            input_values["user_input"] = self.current_user_input
+        return input_values
+
     @property
     def llm_client(self) -> LLMClient:
         """Return the LLM client."""
@@ -313,7 +348,7 @@ class BaseChat(ABC):
         return user_params
 
     async def _build_model_request(self) -> ModelRequest:
-        input_values = await self.generate_input_values()
+        input_values = await self.prepare_input_values()
         # Load history
         self.history_messages = self.current_message.get_history_message()
         self.current_message.start_new_round()

packages/dbgpt-app/src/dbgpt_app/scene/chat_dashboard/chat.py

@@ -95,19 +95,6 @@ class ChatDashboard(BaseChat):
             "supported_chat_type": self.dashboard_template["supported_chart_type"],
         }
 
-        # Mapping variable names: compatible with custom prompt template variable names
-        # Get the input_variables of the current prompt
-        input_variables = []
-        if hasattr(self.prompt_template, "prompt") and hasattr(
-            self.prompt_template.prompt, "input_variables"
-        ):
-            input_variables = self.prompt_template.prompt.input_variables
-        # Compatible with question and user_input
-        if "question" in input_variables:
-            input_values["question"] = self.current_user_input
-        if "user_input" in input_variables:
-            input_values["user_input"] = self.current_user_input
-
         return input_values
 
     def do_action(self, prompt_response):

packages/dbgpt-serve/src/dbgpt_serve/prompt/models/models.py

@@ -139,6 +139,7 @@ class ServeDao(BaseDao[ServeEntity, ServeRequest, ServerResponse]):
             user_code=entity.user_code,
             model=entity.model,
             input_variables=entity.input_variables,
+            response_schema=entity.response_schema,
             prompt_language=entity.prompt_language,
             sys_code=entity.sys_code,
             gmt_created=gmt_created_str,