diff --git a/apps/application/flow/step_node/intent_node/i_intent_node.py b/apps/application/flow/step_node/intent_node/i_intent_node.py
index d22d321c842..e57931da0a0 100644
--- a/apps/application/flow/step_node/intent_node/i_intent_node.py
+++ b/apps/application/flow/step_node/intent_node/i_intent_node.py
@@ -20,12 +20,14 @@ class IntentNodeSerializer(serializers.Serializer):
     model_id_type = serializers.CharField(required=False, default='custom', label=_("Model id type"))
     model_id_reference = serializers.ListField(required=False, child=serializers.CharField(), allow_empty=True,
                                                label=_("Reference Field"))
+    prompt_template = serializers.CharField(required=False, allow_blank=True, allow_null=True, label=_("Prompt template"))
     content_list = serializers.ListField(required=True, label=_("Text content"))
     dialogue_number = serializers.IntegerField(required=True, label=
     _("Number of multi-round conversations"))
     model_params_setting = serializers.DictField(required=False,
                                                  label=_("Model parameter settings"))
     branch = IntentBranchSerializer(many=True)
+    output_reason = serializers.BooleanField(required=False, label=_("Output reason"), default=True)
 
 
 class IIntentNode(INode):
diff --git a/apps/application/flow/step_node/intent_node/impl/base_intent_node.py b/apps/application/flow/step_node/intent_node/impl/base_intent_node.py
index cdf8a769abf..ce51521b5e2 100644
--- a/apps/application/flow/step_node/intent_node/impl/base_intent_node.py
+++ b/apps/application/flow/step_node/intent_node/impl/base_intent_node.py
@@ -12,7 +12,7 @@
 from application.flow.step_node.intent_node.i_intent_node import IIntentNode
 from models_provider.models import Model
 from models_provider.tools import get_model_instance_by_model_workspace_id, get_model_credential
-from .prompt_template import PROMPT_TEMPLATE
+from .prompt_template import DEFAULT_PROMPT_TEMPLATE
 
 
 def get_default_model_params_setting(model_id):
@@ -52,7 +52,7 @@ def save_context(self, details, workflow_manage):
         self.context['branch_id'] = details.get('branch_id')
         self.context['category'] = details.get('category')
 
-    def execute(self, model_id, dialogue_number, history_chat_record, user_input, branch,
+    def execute(self, model_id, prompt_template, dialogue_number, history_chat_record, user_input, branch, output_reason,
                 model_params_setting=None, model_id_type=None, model_id_reference=None, **kwargs) -> NodeResult:
         # 处理引用类型
         if model_id_type == 'reference' and model_id_reference:
@@ -75,18 +75,19 @@ def execute(self, model_id, dialogue_number, history_chat_record, user_input, br
         )
 
         # 获取历史对话
-        history_message = self.get_history_message(history_chat_record, dialogue_number)
+        history_message = self.get_history_message(history_chat_record, dialogue_number) if history_chat_record and dialogue_number > 0 else []
         self.context['history_message'] = history_message
 
         # 保存问题到上下文
         self.context['user_input'] = user_input
 
         # 构建分类提示词
-        prompt = self.build_classification_prompt(user_input, branch)
+        prompt_template = self.workflow_manage.generate_prompt(prompt_template) if prompt_template else None
+        prompt = self.build_classification_prompt(prompt_template, user_input, branch, output_reason)
+        self.context['system'] = prompt
 
         # 生成消息列表
-        system = self.build_system_prompt()
-        message_list = self.generate_message_list(system, prompt, history_message)
+        message_list = self.generate_message_list(prompt, history_message)
         self.context['message_list'] = message_list
 
         # 调用模型进行分类
@@ -104,7 +105,7 @@ def execute(self, model_id, dialogue_number, history_chat_record, user_input, br
                 'history_message': history_message,
                 'user_input': user_input,
                 'branch_id': matched_branch['id'],
-                'reason': self.parse_result_reason(r.content),
+                'reason': self.parse_result_reason(r.content) if output_reason is not False else '',
                 'category': matched_branch.get('content', matched_branch['id'])
             }, {}, _write_context=write_context)
 
@@ -134,11 +135,7 @@ def get_history_message(history_chat_record, dialogue_number):
                 message.content = re.sub('<form_rander>[\d\D]*?<\/form_rander>', '', message.content)
         return history_message
 
-    def build_system_prompt(self) -> str:
-        """构建系统提示词"""
-        return "你是一个专业的意图识别助手，请根据用户输入和意图选项，准确识别用户的真实意图。"
-
-    def build_classification_prompt(self, user_input: str, branch: List[Dict]) -> str:
+    def build_classification_prompt(self, prompt_template: str, user_input: str, branch: List[Dict], output_reason: bool) -> str:
         """构建分类提示词"""
 
         classification_list = []
@@ -160,18 +157,19 @@ def build_classification_prompt(self, user_input: str, branch: List[Dict]) -> st
                 })
                 classification_id += 1
 
-        return PROMPT_TEMPLATE.format(
-            classification_list=classification_list,
-            user_input=user_input
+        # 构建输出JSON结构
+        reason_field = ',\n"reason": ""' if output_reason is not False else ''
+        output_json = f'{{\n"classificationId": 0{reason_field}\n}}'
+
+        return (prompt_template or DEFAULT_PROMPT_TEMPLATE).format(
+            classification_list=json.dumps(classification_list, ensure_ascii=False),
+            user_input=user_input,
+            output_json=output_json
         )
 
-    def generate_message_list(self, system: str, prompt: str, history_message):
+    def generate_message_list(self, prompt: str, history_message):
         """生成消息列表"""
-        if system is None or len(system) == 0:
-            return [*history_message, HumanMessage(self.workflow_manage.generate_prompt(prompt))]
-        else:
-            return [SystemMessage(self.workflow_manage.generate_prompt(system)), *history_message,
-                    HumanMessage(self.workflow_manage.generate_prompt(prompt))]
+        return [*history_message, HumanMessage(self.workflow_manage.generate_prompt(prompt))]
 
     def parse_classification_result(self, result: str, branch: List[Dict]) -> Dict[str, Any]:
         """解析分类结果"""
@@ -191,14 +189,23 @@ def get_branch_by_id(category_id: int):
             return None
 
         try:
-            result_json = json.loads(result)
-            classification_id = result_json.get('classificationId')
+            classification_id = None
+
+            # 如果长度小于5，先尝试解析为数字（增加自由度，在自定义提示词模板时，可提示大模型只输出意图分类的ID值）
+            if len(result) < 5:
+                classification_id = self.to_int(result)
+
+            # 尝试解析为 JSON
+            if classification_id is None:
+                result_json = json.loads(result)
+                classification_id = result_json.get('classificationId')
+
             # 如果是 0 ，返回其他分支
             matched_branch = get_branch_by_id(classification_id)
             if matched_branch:
                 return matched_branch
 
-        except Exception as e:
+        except Exception:
             # json 解析失败，re 提取
             numbers = re.findall(r'"classificationId":\s*(\d+)', result)
             if numbers:
@@ -216,7 +223,7 @@ def parse_result_reason(self, result: str):
         try:
             result_json = json.loads(result)
             return result_json.get('reason', '')
-        except Exception as e:
+        except Exception:
             reason_patterns = [
                 r'"reason":\s*"([^"]*)"',  # 标准格式
                 r'"reason":\s*"([^"]*)',  # 缺少结束引号
@@ -232,6 +239,12 @@ def parse_result_reason(self, result: str):
 
             return ''
 
+    def to_int(self, str):
+        try:
+            return int(str)
+        except ValueError:
+            return None
+
     def find_other_branch(self, branch: List[Dict]) -> Dict[str, Any] | None:
         """查找其他分支"""
         for b in branch:
diff --git a/apps/application/flow/step_node/intent_node/impl/prompt_template.py b/apps/application/flow/step_node/intent_node/impl/prompt_template.py
index 1bcfd61743e..79062022edb 100644
--- a/apps/application/flow/step_node/intent_node/impl/prompt_template.py
+++ b/apps/application/flow/step_node/intent_node/impl/prompt_template.py
@@ -1,6 +1,6 @@
 
 
-PROMPT_TEMPLATE = """# Role
+DEFAULT_PROMPT_TEMPLATE = """# Role
 You are an intention classification expert, good at being able to judge which classification the user's input belongs to.
 
 ## Skills
@@ -20,10 +20,7 @@
 - Strictly ensure that the output is in a valid JSON format.
 - Do not add prefix ```json or suffix ```
 - The answer needs to include the following fields such as:
-{{
-"classificationId": 0,
-"reason": ""
-}}
+{output_json}
 
 ## Limit
 - Please do not reply in text."""
diff --git a/ui/src/components/execution-detail-card/index.vue b/ui/src/components/execution-detail-card/index.vue
index 347ee55ea46..74c645e1bfc 100644
--- a/ui/src/components/execution-detail-card/index.vue
+++ b/ui/src/components/execution-detail-card/index.vue
@@ -194,7 +194,7 @@
                 {{ $t('views.application.form.roleSettings.label') }}
               </h5>
               <div class="p-8-12 border-t-dashed lighter">
-                {{ data.system || '-' }}
+                <pre>{{ data.system || '-' }}</pre>
               </div>
             </div>
             <div class="card-never border-r-6 mt-8" v-if="!isKnowLedge">
@@ -257,7 +257,7 @@
                 {{ $t('views.application.form.roleSettings.label') }}
               </h5>
               <div class="p-8-12 border-t-dashed lighter">
-                {{ data.system || '-' }}
+                <pre>{{ data.system || '-' }}</pre>
               </div>
             </div>
             <div class="card-never border-r-6 mt-8" v-if="data.type !== WorkflowType.Application">
@@ -547,7 +547,7 @@
                 {{ $t('views.application.form.roleSettings.label') }}
               </h5>
               <div class="p-8-12 border-t-dashed lighter">
-                {{ data.system || '-' }}
+                <pre>{{ data.system || '-' }}</pre>
               </div>
             </div>
             <div class="card-never border-r-6 mt-8" v-if="!isKnowLedge">
@@ -633,7 +633,7 @@
                 {{ $t('views.application.form.roleSettings.label') }}
               </h5>
               <div class="p-8-12 border-t-dashed lighter">
-                {{ data.system || '-' }}
+                <pre>{{ data.system || '-' }}</pre>
               </div>
             </div>
             <div class="card-never border-r-6 mt-8" v-if="!isKnowLedge">
diff --git a/ui/src/locales/lang/en-US/views/application.ts b/ui/src/locales/lang/en-US/views/application.ts
index 072779082e7..7bd7272cae2 100644
--- a/ui/src/locales/lang/en-US/views/application.ts
+++ b/ui/src/locales/lang/en-US/views/application.ts
@@ -87,6 +87,11 @@ export default {
          - Please use concise and professional language to answer the user's question.
          `,
     },
+    prompt_template: {
+      label: 'Prompt Template',
+      requiredMessage: 'Please enter Prompt template',
+      tooltip: 'Please pay attention to the placeholders in the template: {classification_list}、{user_input}、{output_json}',
+    },
     historyRecord: {
       label: 'Chat History',
     },
diff --git a/ui/src/locales/lang/en-US/workflow.ts b/ui/src/locales/lang/en-US/workflow.ts
index 74fa49f04b4..7b24d918972 100644
--- a/ui/src/locales/lang/en-US/workflow.ts
+++ b/ui/src/locales/lang/en-US/workflow.ts
@@ -188,7 +188,7 @@ export default {
       result: 'Search Results',
       searchParam: 'Search Parameters',
       select_variable: 'Select Variable',
-      valueMessage: `Value or name `,
+      valueMessage: 'Value or name',
 
       searchQuestion: {
         label: 'Search Question',
@@ -447,9 +447,34 @@ You are a master of problem optimization, adept at accurately inferring user int
       classify: {
         label: 'Intent classify',
       },
+      output_reason: 'Output Reason',
       input: {
         label: 'Input',
       },
+      default_prompt_template: `# Role
+You are an intention classification expert, good at being able to judge which classification the user's input belongs to.
+
+## Skills
+Skill 1: Clearly determine which of the following intention classifications the user's input belongs to.
+Intention classification list:
+{classification_list}
+
+Note:
+- Please determine the match between the user's input content and the Intention classification list content, without judging or categorizing the match with the classification ID.
+- **When classifying, you must give higher weight to the context and intent continuity shown in the historical conversation. Do not rely solely on the literal meaning of the current input; instead, prioritize the most consistent classification with the previous dialogue flow.**
+
+## User Input
+{user_input}
+
+## Reply requirements
+- The answer must be returned in JSON format.
+- Strictly ensure that the output is in a valid JSON format.
+- Do not add prefix \`\`\`json or suffix \`\`\`
+- The answer needs to include the following fields such as:
+{output_json}
+
+## Limit
+- Please do not reply in text.`,
     },
     applicationNode: {
       label: 'Agent Node',
@@ -541,6 +566,7 @@ You are a master of problem optimization, adept at accurately inferring user int
   },
   SystemPromptPlaceholder: 'System Prompt, can reference variables in the system, such as',
   UserPromptPlaceholder: 'User Prompt, can reference variables in the system, such as',
+  PromptTemplatePlaceholder: 'Prompt Template, can reference variables in the system, such as',
   initiator: 'Iniiator',
   abnormalInformation: 'Abnormal Information',
 }
diff --git a/ui/src/locales/lang/zh-CN/views/application.ts b/ui/src/locales/lang/zh-CN/views/application.ts
index 6ebe33d24ce..72e414202ed 100644
--- a/ui/src/locales/lang/zh-CN/views/application.ts
+++ b/ui/src/locales/lang/zh-CN/views/application.ts
@@ -80,6 +80,11 @@ export default {
 回答要求：
 - 请使用中文回答用户问题`,
     },
+    prompt_template: {
+      label: '提示词模板',
+      requiredMessage: '请输入提示词模板',
+      tooltip: '请注意模板中的占位符: {classification_list}、{user_input}、{output_json}',
+    },
     historyRecord: {
       label: '历史聊天记录',
     },
diff --git a/ui/src/locales/lang/zh-CN/workflow.ts b/ui/src/locales/lang/zh-CN/workflow.ts
index 0f069ddeccd..3fa4e620d27 100644
--- a/ui/src/locales/lang/zh-CN/workflow.ts
+++ b/ui/src/locales/lang/zh-CN/workflow.ts
@@ -439,9 +439,34 @@ export default {
       classify: {
         label: '意图分类',
       },
+      output_reason: '输出理由',
       input: {
         label: '输入',
       },
+      default_prompt_template: `# 角色
+你是一位意图分类专家，擅长判断用户输入属于哪个分类。
+
+## 技能
+技能1：明确判断用户输入属于以下哪种意图分类。
+意图分类列表：
+{classification_list}
+
+注：
+- 请判断用户输入内容与意图分类列表内容之间的匹配度，注意不要以分类ID作为评判或归类的依据。
+- **在分类时，必须更加重视历史对话中表现的上下文和意图连贯性。不要仅依赖当前输入的字面意思；相反，应优先考虑与先前对话流程最匹配的分类。**
+
+## 用户输入
+{user_input}
+
+## 回复要求
+- 回复内容必须以JSON格式返回。
+- 严格确保输出为有效的JSON格式。
+- 不要添加前缀 \`\`\`json 或 后缀 \`\`\`
+- 回复内容需要包含以下字段：
+{output_json}
+
+## 限制
+- 请勿以文本形式回复。`,
     },
     applicationNode: {
       label: '智能体节点',
@@ -532,6 +557,7 @@ export default {
   },
   SystemPromptPlaceholder: '系统提示词，可以引用系统中的变量：如',
   UserPromptPlaceholder: '用户提示词，可以引用系统中的变量：如',
+  PromptTemplatePlaceholder: '提示词模板，可以引用系统中的变量：如',
   initiator: '发起人',
   abnormalInformation: '异常信息'
 
diff --git a/ui/src/locales/lang/zh-Hant/views/application.ts b/ui/src/locales/lang/zh-Hant/views/application.ts
index 9a613106305..c2fc674ca07 100644
--- a/ui/src/locales/lang/zh-Hant/views/application.ts
+++ b/ui/src/locales/lang/zh-Hant/views/application.ts
@@ -80,6 +80,11 @@ export default {
 回答要求：
 - 請使用中文回答用戶問題`,
     },
+    prompt_template: {
+      label: '提示詞模板',
+      requiredMessage: '請輸入提示詞模板',
+      tooltip: '請注意模板中的佔位符: {classification_list}、{user_input}、{output_json}',
+    },
     historyRecord: {
       label: '歷史對話紀錄',
     },
diff --git a/ui/src/locales/lang/zh-Hant/workflow.ts b/ui/src/locales/lang/zh-Hant/workflow.ts
index 25d844d783f..3d01130e9b5 100644
--- a/ui/src/locales/lang/zh-Hant/workflow.ts
+++ b/ui/src/locales/lang/zh-Hant/workflow.ts
@@ -439,9 +439,34 @@ export default {
       classify: {
         label: '意圖分類',
       },
+      output_reason: '輸出理由',
       input: {
         label: '輸入',
       },
+      default_prompt_template: `# 角色
+你是一位意圖分類專家，擅長判斷用戶輸入屬於哪個分類。
+
+## 技能
+技能1：明確判斷用戶輸入屬於以下哪種意圖分類。
+意圖分類列表：
+{classification_list}
+
+注：
+- 請判斷用戶輸入內容與意圖分類列表內容之間的匹配度，注意不要以分類ID作爲評判或歸類的依據。
+- **在分類時，必須更加重視歷史對話中表現的上下文和意圖連貫性。不要僅依賴當前輸入的字面意思；相反，應優先考慮與先前對話流程最匹配的分類。**
+
+## 用戶輸入
+{user_input}
+
+## 回覆要求
+- 回覆內容必須以JSON格式返回。
+- 嚴格確保輸出爲有效的JSON格式。
+- 不要添加前綴 \`\`\`json 或 後綴 \`\`\`
+- 回覆內容需要包含以下字段：
+{output_json}
+
+## 限制
+- 請勿以文本形式回覆。`,
     },
     applicationNode: {
       label: '智能體節點',
@@ -526,6 +551,7 @@ export default {
   },
   SystemPromptPlaceholder: '系統提示詞，可以引用系統中的變量：如',
   UserPromptPlaceholder: '用戶提示詞，可以引用系統中的變量：如',
+  PromptTemplatePlaceholder: '提示詞模板，可以引用系統中的變量：如',
   initiator: '發起人',
   abnormalInformation: '異常信息',
 }
diff --git a/ui/src/styles/app.scss b/ui/src/styles/app.scss
index 5d79261d168..364a5f8ca82 100644
--- a/ui/src/styles/app.scss
+++ b/ui/src/styles/app.scss
@@ -18,7 +18,7 @@ html {
   font-size: 100%;
 }
 
-body {
+body, .card-never pre {
   -moz-osx-font-smoothing: grayscale;
   -webkit-font-smoothing: antialiased;
   font-family: 'PingFang SC', AlibabaPuHuiTi !important;
diff --git a/ui/src/workflow/nodes/intent-classify-node/index.ts b/ui/src/workflow/nodes/intent-classify-node/index.ts
index c39167e9294..406a66f5278 100644
--- a/ui/src/workflow/nodes/intent-classify-node/index.ts
+++ b/ui/src/workflow/nodes/intent-classify-node/index.ts
@@ -48,11 +48,11 @@ class IntentModel extends AppNodeModel {
 
         if (branch_condition_list) {
 
-            const FORM_ITEMS_HEIGHT = 397  // 上方表单占用高度
-            
+            const FORM_ITEMS_HEIGHT = 542  // 上方表单占用高度
+
             for (let index = 0; index < branch_condition_list.length; index++) {
                 const element = branch_condition_list[index]
-               
+
                 anchors.push({
                 x: x + width / 2 - 10,
                 y: showNode
diff --git a/ui/src/workflow/nodes/intent-classify-node/index.vue b/ui/src/workflow/nodes/intent-classify-node/index.vue
index 94078687bbc..2ca7af90902 100644
--- a/ui/src/workflow/nodes/intent-classify-node/index.vue
+++ b/ui/src/workflow/nodes/intent-classify-node/index.vue
@@ -76,6 +76,42 @@
             v-model="form_data.model_id_reference"
           />
         </el-form-item>
+        <el-form-item
+          :label="$t('views.application.form.prompt_template.label')"
+          prop="prompt_template"
+          :rules="{
+            required: true,
+            message: $t('views.application.form.prompt_template.requiredMessage'),
+            trigger: 'blur',
+          }"
+        >
+          <template #label>
+            <div class="flex align-center">
+              <div class="mr-4">
+                <span
+                >{{ $t('views.application.form.prompt_template.label')
+                  }}<span class="color-danger">*</span></span
+                >
+              </div>
+              <el-tooltip effect="dark" placement="right" popper-class="max-w-200">
+                <template #content>{{ $t('views.application.form.prompt_template.tooltip', {
+                  classification_list: '{classification_list}',
+                  user_input: '{user_input}',
+                  output_json: '{output_json}',
+                }) }}</template>
+                <AppIcon iconName="app-warning" class="app-warning-icon"></AppIcon>
+              </el-tooltip>
+            </div>
+          </template>
+          <MdEditorMagnify
+            @wheel="wheel"
+            :title="$t('views.application.form.prompt_template.label')"
+            v-model="form_data.prompt_template"
+            style="height: 100px"
+            @submitDialog="submitTemplateDialog"
+            :placeholder="`${t('workflow.PromptTemplatePlaceholder')}{{${t('workflow.nodes.startNode.label')}.question}}`"
+          />
+        </el-form-item>
         <el-form-item
           prop="content_list"
           :label="$t('workflow.nodes.intentNode.input.label')"
@@ -163,6 +199,19 @@
             </div>
           </div>
         </el-form-item>
+        <el-form-item>
+          <div class="flex-between w-full">
+            <div>
+              <span>{{ $t('workflow.nodes.intentNode.output_reason') }}</span>
+            </div>
+            <div>
+              <el-switch
+                size="small"
+                v-model="form_data.output_reason"
+              />
+            </div>
+          </div>
+        </el-form-item>
       </el-form>
     </el-card>
 
@@ -304,10 +353,17 @@ const model_change = (model_id?: string) => {
   }
 }
 
+const default_prompt_template = t('workflow.nodes.intentNode.default_prompt_template', {
+  classification_list: '{classification_list}',
+  user_input: '{user_input}',
+  output_json: '{output_json}',
+})
+
 const form = {
   model_id: '',
   model_id_type: 'custom',
   model_id_reference: [],
+  prompt_template: default_prompt_template,
   branch: [
     {
       id: randomId(),
@@ -322,12 +378,17 @@ const form = {
   ],
   dialogue_number: 1,
   content_list: [],
+  output_reason: true,
 }
 
 function refreshParam(data: any) {
   set(props.nodeModel.properties.node_data, 'model_params_setting', data)
 }
 
+function submitTemplateDialog(val: string) {
+  set(props.nodeModel.properties.node_data, 'prompt_template', val)
+}
+
 const openAIParamSettingDialog = (modelId: string) => {
   if (modelId) {
     AIModeParamSettingDialogRef.value?.open(modelId, id, form_data.value.model_params_setting)
@@ -342,6 +403,9 @@ const form_data = computed({
       if (!props.nodeModel.properties.node_data.model_id_reference) {
         set(props.nodeModel.properties.node_data, 'model_id_reference', [])
       }
+      if (!props.nodeModel.properties.node_data.prompt_template) {
+        set(props.nodeModel.properties.node_data, 'prompt_template', default_prompt_template)
+      }
       return props.nodeModel.properties.node_data
     } else {
       set(props.nodeModel.properties, 'node_data', form)