优化意图识别示例，新增命令行参数解析功能，支持输入输出文件路径和调试模式，增强代码可读性和灵活性。同时更新Dify工具，调整检索信息获取逻辑，确保重排得分信息的正确传递。

rag2_0/demo/intent_recognition_example.py

@@ -17,6 +17,7 @@ import concurrent.futures
 from tqdm import tqdm
 import time
 import sys
+import argparse
 from typing import List, Dict
 # 加载环境变量
 load_dotenv()
@@ -176,6 +177,7 @@ def save_results_to_excel(results, output_file, is_final=False):
 
 # 示例查询
 examples_query = """那储能软件如何操作"""
+examples_query = """博微软件如何新建工程啊"""
 conversation_context=""
 chat_history=[
         {
@@ -199,34 +201,68 @@ previous_slots={
         "software_version": None,
         "operation_steps": None
       }
+
+def parse_arguments():
+    """解析命令行参数"""
+    parser = argparse.ArgumentParser(description='意图识别和问题改写工具')
+    
+    # 添加数据文件路径参数
+    parser.add_argument('--input', '-i', type=str, 
+                        help='输入Excel文件路径，包含待处理的提问数据(第一列)')
+    parser.add_argument('--output', '-o', type=str,
+                        help='输出Excel文件路径，用于保存处理结果')
+    
+    # 添加LLM相关参数
+    parser.add_argument('--model', '-m', type=str,
+                        help='LLM模型名称，默认使用环境变量中的配置')
+    parser.add_argument('--api_base', '-a', type=str,
+                        help='API基础URL，默认使用环境变量中的配置')
+    
+    # 添加处理相关参数
+    parser.add_argument('--max_workers', '-w', type=int, default=20,
+                        help='并发处理的最大线程数，默认为20')
+    parser.add_argument('--debug', '-d', action='store_true',
+                        help='启用调试模式，使用示例查询而非从文件读取')
+    parser.add_argument('--query', '-q', type=str,
+                        help='在调试模式下使用的查询字符串')
+    
+    return parser.parse_args()
+
 def main():
     """
     意图识别和问题改写示例
     """
     
-    # 从环境变量中获取配置
+    # 解析命令行参数
+    args = parse_arguments()
+    
+    # 从环境变量中获取配置，命令行参数优先
     api_key = os.getenv("OPENAI_API_KEY")
-    base_url = os.getenv("OPENAI_API_BASE")
-    model_name = os.getenv("LLM_MODEL_NAME", "gpt-3.5-turbo")
+    base_url = args.api_base if args.api_base else os.getenv("OPENAI_API_BASE")
+    model_name = args.model if args.model else os.getenv("LLM_MODEL_NAME", "gpt-3.5-turbo")
 
     # 初始化意图识别器
     recognizer = IntentRecognizer(api_key=api_key, base_url=base_url, model_name=model_name)
     
     # 读取提问数据
     current_dir = os.path.dirname(os.path.abspath(__file__))
-    data_file = os.path.join(current_dir, "..", "..", "data", "excel", "200条点踩数据测试.xlsx")
-    output_file = os.path.join(current_dir, "..", "..", "data", "excel", "200条点踩数据测试_槽位填充结果.xlsx")
+    data_file = args.input if args.input else os.path.join(current_dir, "..", "..", "data", "excel", "历史提问数据(dislike)_提问明确.xlsx")
+    output_file = args.output if args.output else os.path.join(current_dir, "..", "..", "data", "excel", "历史提问数据(dislike)_槽位(分类)填充结果.xlsx")
+    
+    # 检测是否为调试模式
+    is_debug = args.debug or (hasattr(sys, 'gettrace') and sys.gettrace() is not None)
     
-    # 检测是否为调试模式，调试模式下使用examples_query，否则从Excel读取
-    is_debug = hasattr(sys, 'gettrace') and sys.gettrace() is not None
-    # is_debug = False
     if is_debug:
-        examples = examples_query.strip().split("\n")
+        # 如果提供了查询参数，使用它；否则使用默认示例
+        if args.query:
+            examples = [args.query]
+        else:
+            examples = examples_query.strip().split("\n")
     else:
         examples = load_questions_from_excel(data_file)
     
     if not is_debug:
-        max_workers = 20  # 减少并发数以避免API限制
+        max_workers = args.max_workers
         logging.info(f"共有 {len(examples)} 个问题需要处理，使用 {max_workers} 个并发线程")
         
         # 创建一个与输入顺序相同的结果列表
@@ -262,8 +298,8 @@ def main():
         for idx, query in enumerate(examples):
             if query.strip() == "":
                 continue
-            process_query(recognizer, query, conversation_context, chat_history, previous_slots)
-            # print(json.dumps(process_query(recognizer, query), ensure_ascii=False, indent=2))
+            # process_query(recognizer, query, conversation_context, chat_history, previous_slots)
+            print(json.dumps(process_query(recognizer, query), ensure_ascii=False, indent=2))
 
 def setup_logging():
     # 配置日志输出到控制台

rag2_0/dify/dify_tool.py

@@ -318,7 +318,7 @@ content: "{content}"
         except Exception as e:
             return -1
     
-    def get_retrieve_info(self, query: str, outputs: dict) -> tuple:
+    def get_retrieve_info(self, query: str, outputs: dict, reranker_sorce_info:list) -> tuple:
         """
         获取检索信息并计算分数
         
@@ -333,20 +333,21 @@ content: "{content}"
         min_score = 10
         total_score = 0
         valid_scores = 0
-        retrieve_content = []
+        retrieve_title = []
         
         # 使用线程池并发计算分数
         with ThreadPoolExecutor() as executor:
             # 创建任务列表
             future_to_content = {}
-            for result in outputs["result"]:
-                content = result["content"].strip()
+            for result in outputs:
+                content = result["segment_content"].strip()
+                segment_id = result["segment_id"].strip()
                 future = executor.submit(self.calculate_score, query=query, content=content)
-                future_to_content[future] = content
+                future_to_content[future] = (content, segment_id)
             
             # 收集结果
             for future in as_completed(future_to_content):
-                content = future_to_content[future]
+                content, segment_id = future_to_content[future]
                 score = future.result()
                 content_title = content.split("\n")[0]
                 
@@ -357,10 +358,11 @@ content: "{content}"
                     valid_scores += 1
                 
                 if content_title:
-                    retrieve_content.append(content_title + f"--相关性得分({score}分)")
+                    current_score = next((cur_source_info["score"] for cur_source_info in reranker_sorce_info if cur_source_info["segment_id"] == segment_id), None)
+                    retrieve_title.append(content_title + f"--LLM得分({score}分)--重排得分({current_score:.2f}分)")
                 
         avg_score = total_score / valid_scores if valid_scores > 0 else 0
-        return retrieve_content, max_score, min_score, avg_score
+        return retrieve_title, max_score, min_score, avg_score
 
 
 class NewWorkflowChat(BaseWorkflowChat):
@@ -395,7 +397,6 @@ class NewWorkflowChat(BaseWorkflowChat):
             "新问题分类": workflow_info["问题分类"],
             "槽点信息": workflow_info["槽点信息"],
             "新检索词条": workflow_info["检索词条"],
-            "检索内容": workflow_info["检索内容"],
             "message_id":message_id
         }
         
@@ -421,14 +422,23 @@ class NewWorkflowChat(BaseWorkflowChat):
         vertical_classification = ""
         sub_classification = ""
         slot_info = ""
-        
+        reranker_sorce=[]
         try:
+            # 先取出重排得分
             message_info = DifyTool.get_message_debug_info_by_id(message_id=message_id)
             for workflow_node in message_info["workflow_node_executions_info"]:
-                if workflow_node["title"] == "知识检索结果后处理":
-                    outputs = json.loads(workflow_node["outputs"])  
-                    retrieve_title, max_score, min_score, avg_score = self.get_retrieve_info(query=query, outputs=outputs)
-                    retrieve_content = outputs["result"]
+                if workflow_node["title"] == "软件知识检索聚合":
+                    retrieve_outputs = json.loads(workflow_node["inputs"])["result"]
+                    reranker_sorce = [{"score":result["metadata"]["score"], "segment_id":result["metadata"]["segment_id"]} for result in retrieve_outputs]
+
+
+            for workflow_node in message_info["workflow_node_executions_info"]:
+                if workflow_node["title"] == "软件知识检索聚合":
+                    retrieve_outputs = json.loads(workflow_node["inputs"])["result"]
+                    reranker_sorce = [{"score":result["metadata"]["score"], "segment_id":result["metadata"]["segment_id"]} for result in retrieve_outputs]
+                elif workflow_node["title"] == "提取处理后的知识":
+                    outputs = json.loads(workflow_node["outputs"])["knowledge_list"]
+                    retrieve_title, max_score, min_score, avg_score = self.get_retrieve_info(query=query, outputs=outputs, reranker_sorce_info=reranker_sorce)
                 elif workflow_node["title"] == "问题优化结果解析":
                     outputs = json.loads(workflow_node["outputs"])  
                     rewrite_query = outputs["optimize_query"]
@@ -439,11 +449,18 @@ class NewWorkflowChat(BaseWorkflowChat):
                     slot_info = json.dumps(json_result["slot_filling"], ensure_ascii=False, indent=2)
         except Exception as e:
             raise e
-            
+        
+        retrieve_content = ""
+        if len(reranker_sorce)==0:
+            retrieve_content="未检索知识库"
+        elif len(reranker_sorce) > 0 and len(retrieve_title)==0:
+            retrieve_content = "知识与提问不相关，被丢弃"
+        else:
+            retrieve_content = "\n".join(retrieve_title)
+
         return {
             "问题改写": rewrite_query,
-            "检索词条": "\n".join(retrieve_title) if retrieve_title else "未检索知识库",
-            "检索内容": retrieve_content,
+            "检索词条": retrieve_content,
             "问题分类": f"{vertical_classification} - {sub_classification}",
             "槽点信息": slot_info,
 
@@ -479,7 +496,6 @@ class OldWorkFlowChat(BaseWorkflowChat):
             "旧流程答案": answer,
             "旧问题改写": workflow_info["问题改写"],
             "旧检索词条": workflow_info["检索词条"],
-            "检索内容": workflow_info["检索内容"],
             "message_id":message_id
         }
         
@@ -519,7 +535,6 @@ class OldWorkFlowChat(BaseWorkflowChat):
         return {
             "问题改写": rewrite_query,
             "检索词条": "\n".join(retrieve_title) if retrieve_title else "未检索知识库",
-            "检索内容": retrieve_content,
         } 
 
 if __name__ == "__main__":

rag2_0/dify/test_dify_chatapi.py

@@ -411,7 +411,6 @@ content: "{content}"
         return {
             "问题改写": rewrite_query,
             "检索词条": "\n".join(retrieve_title) if retrieve_title else "未检索知识库",
-            "检索内容": retrieve_content,
             "问题分类": f"{vertical_classification} - {sub_classification}",
             "槽点信息": slot_info
         }
@@ -451,7 +450,6 @@ content: "{content}"
         return {
             "问题改写": rewrite_query,
             "检索词条": "\n".join(retrieve_title) if retrieve_title else "未检索知识库",
-            "检索内容": retrieve_content,
         }
 
     def get_retrieve_title_similarity(self, old_retrieve_content:list[dict], new_retrieve_content:list[dict]) -> str:
@@ -589,9 +587,7 @@ content: "{content}"
             
         if judge_result is None:
             judge_result = ""
-        
-        # retrieve_title_score = self.get_retrieve_title_similarity(old_retrieve_content=old_workflow_info["检索内容"], new_retrieve_content=new_workflow_info["检索内容"])
-        
+           
         # 返回结果
         return {
             "问题": query,

rag2_0/dify/test_workorder.py

@@ -0,0 +1,101 @@
+import pandas as pd
+import random
+import math
+
+work_order_excel="data/excel/6万工单记录.xlsx"
+
+soft_row_data={
+    "博微配网计价通D3":{"基本功能":[], "高级功能":[]},
+    "储能C1软件":{"基本功能":[], "高级功能":[]},
+    "西藏计价通Z1":{"基本功能":[], "高级功能":[]},
+    "技改检修工程计价通T1":{"基本功能":[], "高级功能":[]},
+    "检修清单计价通T1":{"基本功能":[], "高级功能":[]},
+    "电力建设计价通软件":{"基本功能":[], "高级功能":[]},
+}
+
+df = pd.read_excel(work_order_excel)
+
+for idx, row in df.iterrows():
+    if pd.isna(row["产品线"]):
+        continue
+
+    if "博微配网计价通D3" in row["产品线"]:
+        soft_row_data["博微配网计价通D3"][row["问题类型"]].append((idx, row))
+    elif "博微电力建设计价通软件" in row["产品线"]:
+        soft_row_data["电力建设计价通软件"][row["问题类型"]].append((idx, row))
+    elif "新能源系列" in row["产品线"] and "博微新型储能电站建设计价通C1软件" in row["产品名称"]:
+        soft_row_data["储能C1软件"][row["问题类型"]].append((idx, row))
+    elif "博微西藏计价通Z1" in row["产品线"]:
+        soft_row_data["西藏计价通Z1"][row["问题类型"]].append((idx, row))
+    elif "博微技改检修计价通T1软件" in row["产品线"] and "技改检修计价通T1软件-概预算" in row["产品名称"]:
+        soft_row_data["技改检修工程计价通T1"][row["问题类型"]].append((idx, row))
+    elif "博微技改检修计价通T1软件" in row["产品线"] and "技改检修计价通T1软件-清单" in row["产品名称"]:
+        soft_row_data["检修清单计价通T1"][row["问题类型"]].append((idx, row))
+
+# 计算每个软件和功能类型的数据量
+total_count = 0
+counts = {}
+for software, types in soft_row_data.items():
+    counts[software] = {}
+    for type_name, rows in types.items():
+        counts[software][type_name] = len(rows)
+        total_count += len(rows)
+
+print(f"原始数据总量: {total_count}条")
+for software, types in counts.items():
+    print(f"{software}: 基本功能 {types['基本功能']}条, 高级功能 {types['高级功能']}条")
+
+# 计算均衡提取的数量
+total_target = 2000
+categories_count = sum(len(types) for types in soft_row_data.values())
+per_category_target = math.ceil(total_target / categories_count)
+
+# 均衡提取数据
+balanced_data = []
+extracted_counts = {}
+extracted_indices = set()  # 使用集合存储已提取数据的索引
+
+for software, types in soft_row_data.items():
+    extracted_counts[software] = {}
+    
+    for type_name, rows in types.items():
+        # 如果数据量不足，全部提取；否则随机抽取目标数量
+        if len(rows) <= per_category_target:
+            extracted = rows
+        else:
+            extracted = random.sample(rows, per_category_target)
+        
+        extracted_counts[software][type_name] = len(extracted)
+        for idx, row in extracted:
+            extracted_indices.add(idx)  # 记录已提取数据的索引
+            balanced_data.append(row)
+
+# 数据量不足2000时，从剩余数据中补充
+remaining_target = total_target - len(balanced_data)
+if remaining_target > 0:
+    # 收集所有未被选中的数据
+    remaining_data = []
+    for software, types in soft_row_data.items():
+        for type_name, rows in types.items():
+            # 添加未被选中的数据
+            for idx, row in rows:
+                if idx not in extracted_indices:
+                    remaining_data.append(row)
+    
+    # 如果剩余数据足够，随机抽取补充
+    if len(remaining_data) >= remaining_target:
+        additional_data = random.sample(remaining_data, remaining_target)
+    else:
+        additional_data = remaining_data
+    
+    balanced_data.extend(additional_data)
+
+# 输出结果
+print(f"\n均衡提取后数据总量: {len(balanced_data)}条")
+for software, types in extracted_counts.items():
+    print(f"{software}: 基本功能 {types['基本功能']}条, 高级功能 {types['高级功能']}条")
+
+# 将均衡提取的数据转换为DataFrame并保存
+balanced_df = pd.DataFrame(balanced_data)
+balanced_df.to_excel("data/excel/均衡提取2000条工单.xlsx", index=False)
+print(f"\n已将均衡提取的{len(balanced_data)}条数据保存至'data/excel/均衡提取2000条工单.xlsx'")

rag2_0/intent_recognition/DataModels.py

@@ -160,7 +160,7 @@ class SoftwareFunctionSlots(SlotBase):
             self.project_type="单工程"
         missing_slots = {}
         if not self.software_name:
-            missing_slots["software_name"] = f"{SoftwareFunctionSlots.model_fields['software_name'].description}，可选值：{', '.join([name.value for name in SoftwareName if name not in [SoftwareName.UNKNOWN, SoftwareName.ALIASES]])}"
+            missing_slots["software_name"] = f"{SoftwareFunctionSlots.model_fields['software_name'].description}，支持的软件：{', '.join([name.value for name in SoftwareName if name not in [SoftwareName.UNKNOWN, SoftwareName.ALIASES]])}"
         if not self.function_name:
             missing_slots["function_name"] = SoftwareFunctionSlots.model_fields["function_name"].description
         if not self.operation:
@@ -181,7 +181,7 @@ class SoftwareTroubleShootingSlots(SlotBase):
         """检查必填槽位是否都存在"""
         missing_slots = {}
         if not self.software_name:
-            missing_slots["software_name"] = f"{SoftwareTroubleShootingSlots.model_fields['software_name'].description}，可选值：{', '.join([name.value for name in SoftwareName if name not in [SoftwareName.UNKNOWN, SoftwareName.ALIASES]])}"
+            missing_slots["software_name"] = f"{SoftwareTroubleShootingSlots.model_fields['software_name'].description}，支持的软件：{', '.join([name.value for name in SoftwareName if name not in [SoftwareName.UNKNOWN, SoftwareName.ALIASES]])}"
         if not self.function_name:
             missing_slots["function_name"] = SoftwareTroubleShootingSlots.model_fields["function_name"].description
         if not self.error_message:
@@ -191,7 +191,7 @@ class SoftwareTroubleShootingSlots(SlotBase):
 # 2. 业务问题
 # 2.1 专业咨询
 class ProfessionalConsultingSlots(SlotBase):
-    scene_subject: str = Field(default="", description="场景主体")
+    scene_subject: str = Field(default="", description="业务主体。即询问的业务对象(规范、标准、费用等)")
     business_scene: str = Field(default="", description="业务场景描述")
     software_name: Optional[str] = Field(default="", description="软件名称")
     
@@ -266,7 +266,6 @@ class InstallationDownloadSlots(SlotBase):
         missing_slots = {}
         if not self.software_name and not self.file_name:
             missing_slots["software_name"] = f"{InstallationDownloadSlots.model_fields['software_name'].description},"
-            f"可选值：{', '.join([name.value for name in SoftwareName if name not in [SoftwareName.UNKNOWN, SoftwareName.ALIASES]])}"
             missing_slots["file_name"] = InstallationDownloadSlots.model_fields["file_name"].description
         if not self.operation_stage:
             missing_slots["operation_stage"] = InstallationDownloadSlots.model_fields["operation_stage"].description

rag2_0/intent_recognition/IntentRecognition.py

@@ -304,8 +304,8 @@ class IntentRecognizer:
 
         rewrite_start_time = time.time()
         # 准备问题改写提示
-        # terms_dict = [term.model_dump(exclude={"description"}) for term in keywords.terms] 
-        terms_dict = [term.model_dump() for term in keywords.terms] 
+        terms_dict = [term.model_dump(exclude={"description"}) for term in keywords.terms] 
+        # terms_dict = [term.model_dump() for term in keywords.terms] 
         keywords_str = json.dumps(terms_dict, ensure_ascii=False)
         query_rewrite_parser = PydanticOutputParser(pydantic_object=QueryRewrite)
         # formatted_prompt = query_rewrite_prompt.format(query=query, 
@@ -401,27 +401,27 @@ class IntentRecognizer:
         )
         
         # 步骤3: 进行意图识别和槽位填充
-        result = self._process_intent_and_slot(rewrite.rewrite, conversation_context, chat_history, previous_slots)
-        result.update({"keywords": keywords_terms.model_dump(),
-                              "rewrite": rewrite.model_dump(),
-                              "query_keys": query_keys})
-        return result
-        # # 步骤3: 进行意图分类
-        # classification = self._classify_intent(query)
+        # result = self._process_intent_and_slot(rewrite.rewrite, conversation_context, chat_history, previous_slots)
+        # result.update({"keywords": keywords_terms.model_dump(),
+        #                       "rewrite": rewrite.model_dump(),
+        #                       "query_keys": query_keys})
+        # return result
+        # 步骤3: 进行意图分类
+        classification = self._classify_intent(rewrite.rewrite, conversation_context, chat_history, previous_slots)
 
-        # # 步骤4: 进行槽位填充   
-        # # 如果是有效分类，进行槽位填充
-        # slot_filling_result = {}
-        # if classification.vertical_classification not in ["其他", "闲聊"] and classification.sub_classification not in ["其他", "闲聊"]:
-        #     slot_filling_result = self._fill_slots(rewrite.rewrite, classification)
+        # 步骤4: 进行槽位填充   
+        # 如果是有效分类，进行槽位填充
+        slot_filling_result = {}
+        if classification.vertical_classification not in ["其他", "闲聊"] and classification.sub_classification not in ["其他", "闲聊"]:
+            slot_filling_result = self._fill_slots(rewrite.rewrite, classification, conversation_context, chat_history, previous_slots)
         
-        # return {
-        #     "classification": classification.model_dump(),
-        #     "keywords": keywords_terms.model_dump(),
-        #     "rewrite": rewrite.model_dump(),
-        #     "query_keys": query_keys,
-        #     "slot_filling": slot_filling_result
-        # }
+        return {
+            "classification": classification.model_dump(),
+            "keywords": keywords_terms.model_dump(),
+            "rewrite": rewrite.model_dump(),
+            "query_keys": query_keys,
+            "slot_filling": slot_filling_result
+        }
 
     
     def _fill_slots(self, query: str, classification: Classification, conversation_context: str = "", 

rag2_0/intent_recognition/Multi_PromptTemplates.py

@@ -127,11 +127,13 @@ query_rewrite_prompt_pro_old="""
 query_rewrite_prompt_pro="""
 # 电力造价问答优化工程师（精简版）
 **角色**：基于历史对话和术语库重构问题，提升知识库检索准确率。
+最高准则：保持问题核心意图，但允许在指代消除、背景继承下添加隐含功能词。但重构后的问题，所有引入的主体背景等均要来源于历史对话、聊天背景或术语库，不得凭空捏造未提及的内容。
 
 ## 核心原则
-1. 语义保真 → 保持问题核心意图
-2. 术语规范 → 同义词转标准词并【】标记
-3. 背景继承 → 补充历史对话的隐含信息
+1. **指代消除 → 当指示代词（"那"/"这"）出现时，强制继承历史对话的最新核心主题（如功能或任务），并应用到当前主体。**  
+2. 背景继承 → 补充历史对话和聊天背景中的隐含信息（包括主题和功能）。  
+4. 术语规范 → 同义词转标准词并【】标记。提问中的同义词(synonymous)替换为标准词(name)   
+5. 语义保真 → 保持问题核心意图，但允许在指代消除、背景继承下添加隐含功能词。
 
 ## 处理流程
 ### 一、输入解析
@@ -155,37 +157,30 @@ query_rewrite_prompt_pro="""
 ### 二、重构决策树
 ```mermaid
 graph TD
-    A[输入问题] --> B{{匹配关键词或上下文？}}
-    B -- 是 --> C[执行重构]
-    B -- 否 --> D[直接输出原始问题]
-    C --> E[补充缺失背景]
-    E --> F[同义词替换+【】标记]
-    F --> G[保留原生专业术语]
+    A[输入问题] --> B{{包含指示代词？}}
+    B -- 是 --> C[提取历史最新主题]
+    C --> D{{主题是否明确？}}
+    D -- 是 --> E[继承主题到当前问题]
+    E --> F[执行重构]
+    D -- 否 --> F
+    F --> G[补充缺失背景]
+    G --> H[同义词替换+【】标记]
+    H --> I[保留原生专业术语]
+    B -- 否 --> I
 ```
 
 ### 三、重构优先级
-1. **背景补充**
-   - 历史对话中确定的背景信息需要保留（例："这软件"→"【配网工程计价通D3软件】"）
- 
-2. **术语处理**
-   - 同义词转标准词 → 将提问中的同义词(synonymous)替换为标准词(name)
-   - 存在即标记 → 【计算式】
-
-3. **结构优化**
-   - 保持原问题的5W2H特征,确保问题意图不发生改变。
-   - 明确指代关系（"该功能"→"【批量导入】功能"）
+1. **指代消除 → 当指示代词出现时，优先继承历史对话的核心主题（如功能词），并替换当前问题的动词部分。**  
+2. 背景继承 → 历史对话中确定的背景信息需要保留。  
+3. 术语处理 → 同义词转标准词 + 【】标记。
+4. 同义词转标准词 → 将提问中的同义词(synonymous)替换为标准词(name) 
+4. 结构优化 → 保持原问题的5W2H特征，指代消除、背景继承下允许微调意图。
 
 ## 输出规范
 {output_format}
 
-## 典型案例
-| 场景                | 输入问题                          | 输出结果                                 |
-|---------------------|-----------------------------------|------------------------------------------|
-| 强上下文关联        | “怎么升级旧版工程”               | {{"rewrite":"【西藏Z1】如何执行【老版本定额升级】？"}} |
-| 弱术语匹配          | “界面文字太小怎么办”             | 原样输出                                |
-| 代词+背景继承       | “这个定额如何导入”               | {{"rewrite":"【山东定额】如何执行【批量导入定额】？"}}|
-
 ## 质量自检
+- [] **主题是否合理继承？**（当有代词时，历史主题必须注入）
 - [] 核心诉求是否保留？
 - [] 背景信息是否合理补充？
 - [] 术语标记是否完整【】？

rag2_0/tool/APIKeyManager.py

@@ -19,7 +19,6 @@ import requests
 
 API_KEY_LIST=[
 "sk-kvgfuqeqvpmfsccykyoohheshclcrtvjlnewratvrjpkpbkc",
-"sk-zhnbqnpuumuuvegnvbgoggxafpukbzchpgrugpkobiwkzsar",
 "sk-kzhxlqvqcxlnbdgnpalqnzumkmspepkttkgbophnkqanainw",
 "sk-bzttugqtlskrvguvhckwamdssvgmgnrqpsialpdbskfsyyak",
 "sk-tovmogiablsoeabwgqyvevpcfichyjpuzqdymmvksspdrtqt",