feat: 更好的对话管理

astrbot/core/conversation_mgr.py

@@ -0,0 +1,118 @@
+import uuid
+import json
+import asyncio
+from astrbot.core import sp
+from typing import Dict, List
+from astrbot.core.db import BaseDatabase
+from astrbot.core.db.po import Conversation
+
+class ConversationManager():
+    '''负责管理会话与 LLM 的对话，某个会话当前正在用哪个对话。'''
+    def __init__(self, db_helper: BaseDatabase):
+        self.session_conversations: Dict[str, str] = sp.get("session_conversation", {})
+        self.db = db_helper
+        self.save_interval = 60  # 每 60 秒保存一次
+        self._start_periodic_save()
+                
+    def _start_periodic_save(self):
+        asyncio.create_task(self._periodic_save())
+
+    async def _periodic_save(self):
+        while True:
+            await asyncio.sleep(self.save_interval)
+            self._save_to_storage()
+
+    def _save_to_storage(self):
+        sp.put("session_conversation", self.session_conversations)
+
+    async def new_conversation(self, unified_msg_origin: str) -> str:
+        '''新建对话，并将当前会话的对话转移到新对话'''
+        conversation_id = str(uuid.uuid4())
+        self.db.new_conversation(
+            user_id=unified_msg_origin,
+            cid=conversation_id
+        )
+        self.session_conversations[unified_msg_origin] = conversation_id
+        sp.put("session_conversation", self.session_conversations)
+        return conversation_id
+    
+    async def switch_conversation(self, unified_msg_origin: str, conversation_id: str):
+        '''切换会话的对话'''
+        self.session_conversations[unified_msg_origin] = conversation_id
+        sp.put("session_conversation", self.session_conversations)
+        
+    async def delete_conversation(self, unified_msg_origin: str, conversation_id: str=None):
+        '''删除会话的对话，当 conversation_id 为 None 时删除会话当前的对话'''
+        conversation_id = self.session_conversations.get(unified_msg_origin)
+        if conversation_id:
+            self.db.delete_conversation(
+                user_id=unified_msg_origin,
+                cid=conversation_id
+            )
+            del self.session_conversations[unified_msg_origin]
+            
+    async def get_curr_conversation_id(self, unified_msg_origin: str) -> str:
+        '''获取会话当前的对话 ID'''
+        return self.session_conversations.get(unified_msg_origin, None)
+    
+    async def get_conversation(self, unified_msg_origin: str, conversation_id: str) -> Conversation:
+        '''获取会话的对话'''
+        return self.db.get_conversation_by_user_id(unified_msg_origin, conversation_id)
+    
+    async def get_conversations(self, unified_msg_origin: str) -> List[Conversation]:
+        '''获取会话的所有对话'''
+        return self.db.get_conversations(unified_msg_origin)
+    
+    async def update_conversation(self, unified_msg_origin: str, conversation_id: str, history: List[Dict]):
+        '''更新会话的对话'''
+        if conversation_id:
+            self.db.update_conversation(
+                user_id=unified_msg_origin,
+                cid=conversation_id,
+                history=json.dumps(history)
+            )
+            
+    async def update_conversation_title(self, unified_msg_origin: str, title: str):
+        '''更新会话的对话标题'''
+        conversation_id = self.session_conversations.get(unified_msg_origin)
+        if conversation_id:
+            self.db.update_conversation_title(
+                user_id=unified_msg_origin,
+                cid=conversation_id,
+                title=title
+            )
+    
+    async def update_conversation_persona_id(self, unified_msg_origin: str, persona_id: str):
+        '''更新会话的对话 Persona ID'''
+        conversation_id = self.session_conversations.get(unified_msg_origin)
+        if conversation_id:
+            self.db.update_conversation_persona_id(
+                user_id=unified_msg_origin,
+                cid=conversation_id,
+                persona_id=persona_id
+            )
+            
+    async def get_human_readable_context(self, unified_msg_origin, conversation_id, page=1, page_size=10):
+        conversation = await self.get_conversation(unified_msg_origin, conversation_id)
+        history = json.loads(conversation.history)
+
+        contexts = []
+        temp_contexts = []
+        for record in history:
+            if record['role'] == "user":
+                temp_contexts.append(f"User: {record['content']}")
+            elif record['role'] == "assistant":
+                temp_contexts.append(f"Assistant: {record['content']}")
+                contexts.insert(0, temp_contexts)
+                temp_contexts = []
+
+        # 展平 contexts 列表
+        contexts = [item for sublist in contexts for item in sublist]
+
+        # 计算分页
+        paged_contexts = contexts[(page-1)*page_size:page*page_size]
+        total_pages = len(contexts) // page_size
+        if len(contexts) % page_size != 0:
+            total_pages += 1
+        
+        return paged_contexts, total_pages

astrbot/core/core_lifecycle.py

@@ -18,7 +18,7 @@ from astrbot.core.updator import AstrBotUpdator
 from astrbot.core import logger
 from astrbot.core.config.default import VERSION
 from astrbot.core.rag.knowledge_db_mgr import KnowledgeDBManager
-
+from astrbot.core.conversation_mgr import ConversationManager
 class AstrBotCoreLifecycle:
     def __init__(self, log_broker: LogBroker, db: BaseDatabase):
         self.log_broker = log_broker
@@ -43,12 +43,15 @@ class AstrBotCoreLifecycle:
         
         self.knowledge_db_manager = KnowledgeDBManager(self.astrbot_config)
         
+        self.conversation_manager = ConversationManager(self.db)
+        
         self.star_context = Context(
             self.event_queue, 
             self.astrbot_config, 
             self.db,
             self.provider_manager,
             self.platform_manager,
+            self.conversation_manager,
             self.knowledge_db_manager
         )
         self.plugin_manager = PluginManager(self.star_context, self.astrbot_config)

astrbot/core/db/__init__.py

@@ -1,7 +1,7 @@
 import abc
 from dataclasses import dataclass
 from typing import List
-from astrbot.core.db.po import Stats, LLMHistory, ATRIVision, WebChatConversation
+from astrbot.core.db.po import Stats, LLMHistory, ATRIVision, Conversation
 
 @dataclass
 class BaseDatabase(abc.ABC):
@@ -79,25 +79,35 @@ class BaseDatabase(abc.ABC):
         raise NotImplementedError
     
     @abc.abstractmethod
-    def get_webchat_conversation_by_user_id(self, user_id: str, cid: str) -> WebChatConversation:
-        '''通过 user_id 和 cid 获取 WebChatConversation'''
+    def get_conversation_by_user_id(self, user_id: str, cid: str) -> Conversation:
+        '''通过 user_id 和 cid 获取 Conversation'''
         raise NotImplementedError
     
     @abc.abstractmethod
-    def webchat_new_conversation(self, user_id: str, cid: str):
-        '''新建 WebChatConversation'''
+    def new_conversation(self, user_id: str, cid: str):
+        '''新建 Conversation'''
         raise NotImplementedError
     
     @abc.abstractmethod
-    def get_webchat_conversations(self, user_id: str) -> List[WebChatConversation]:
+    def get_conversations(self, user_id: str) -> List[Conversation]:
         raise NotImplementedError
 
     @abc.abstractmethod
-    def update_webchat_conversation(self, user_id: str, cid: str, history: str):
-        '''更新 WebChatConversation'''
+    def update_conversation(self, user_id: str, cid: str, history: str):
+        '''更新 Conversation'''
         raise NotImplementedError
     
     @abc.abstractmethod
-    def delete_webchat_conversation(self, user_id: str, cid: str):
-        '''删除 WebChatConversation'''
+    def delete_conversation(self, user_id: str, cid: str):
+        '''删除 Conversation'''
+        raise NotImplementedError
+    
+    @abc.abstractmethod
+    def update_conversation_title(self, user_id: str, cid: str, title: str):
+        '''更新 Conversation 标题'''
+        raise NotImplementedError
+    
+    @abc.abstractmethod
+    def update_conversation_persona_id(self, user_id: str, cid: str, persona_id: str):
+        '''更新 Conversation Persona ID'''
         raise NotImplementedError

astrbot/core/db/po.py

@@ -33,16 +33,16 @@ class Stats():
     command: List[Command] = field(default_factory=list)
     llm: List[Provider] = field(default_factory=list)
     
-'''LLM 聊天时持久化的信息'''
-
 @dataclass
 class LLMHistory():
+    '''LLM 聊天时持久化的信息'''
     provider_type: str
     session_id: str
     content: str
     
 @dataclass
 class ATRIVision():
+    '''Deprecated'''
     id: str
     url_or_path: str
     caption: str
@@ -53,13 +53,18 @@ class ATRIVision():
     sender_nickname: str
     timestamp: int = -1
     
-    
-
 @dataclass
-class WebChatConversation():
+class Conversation():
+    '''LLM 对话存储
+    
+    对于网页聊天，history 存储了包括指令、回复、图片等在内的所有消息。
+    对于其他平台的聊天，不存储非 LLM 的回复（因为考虑到已经存储在各自的平台上）。
+    '''
     user_id: str
     cid: str
     history: str = ""
+    '''字符串格式的列表。'''
     created_at: int = 0
     updated_at: int = 0
-    
+    title: str = ""
+    persona_id: str = ""

astrbot/core/db/sqlite.py

@@ -6,7 +6,7 @@ from astrbot.core.db.po import (
     Stats,
     LLMHistory,
     ATRIVision,
-    WebChatConversation
+    Conversation
 )
 from . import BaseDatabase
 from typing import Tuple
@@ -25,6 +25,37 @@ class SQLiteDatabase(BaseDatabase):
         c = self.conn.cursor()
         c.executescript(sql)
         self.conn.commit()
+        
+        # 检查 webchat_conversation 的 title 字段是否存在
+        c.execute(
+            '''
+            PRAGMA table_info(webchat_conversation)
+            '''
+        )
+        res = c.fetchall()
+        has_title = False
+        has_persona_id = False
+        for row in res:
+            if row[1] == "title":
+                has_title = True
+            if row[1] == "persona_id":
+                has_persona_id = True
+        if not has_title:
+            c.execute(
+                '''
+                ALTER TABLE webchat_conversation ADD COLUMN title TEXT;
+                '''
+            )
+            self.conn.commit()
+        if not has_persona_id:
+            c.execute(
+                '''
+                ALTER TABLE webchat_conversation ADD COLUMN persona_id TEXT;
+                '''
+            )
+            self.conn.commit()
+            
+        c.close()
     
     def _get_conn(self, db_path: str) -> sqlite3.Connection:
         conn = sqlite3.connect(self.db_path)
@@ -202,7 +233,7 @@ class SQLiteDatabase(BaseDatabase):
         return Stats(platform, [], [])
     
     
-    def get_webchat_conversation_by_user_id(self, user_id: str, cid: str) -> WebChatConversation:
+    def get_conversation_by_user_id(self, user_id: str, cid: str) -> Conversation:
         try:
             c = self.conn.cursor()
         except sqlite3.ProgrammingError:
@@ -216,9 +247,9 @@ class SQLiteDatabase(BaseDatabase):
         
         res = c.fetchone()
         c.close()
-        return WebChatConversation(*res)
+        return Conversation(*res)
     
-    def webchat_new_conversation(self, user_id: str, cid: str):
+    def new_conversation(self, user_id: str, cid: str):
         history = "[]"
         updated_at = int(time.time())
         created_at = updated_at
@@ -228,7 +259,7 @@ class SQLiteDatabase(BaseDatabase):
             ''', (user_id, cid, history, updated_at, created_at)
         )
         
-    def get_webchat_conversations(self, user_id: str) -> Tuple:
+    def get_conversations(self, user_id: str) -> Tuple:
         try:
             c = self.conn.cursor()
         except sqlite3.ProgrammingError:
@@ -236,7 +267,7 @@ class SQLiteDatabase(BaseDatabase):
             
         c.execute(
             '''
-            SELECT cid, created_at, updated_at FROM webchat_conversation WHERE user_id = ? ORDER BY updated_at DESC
+            SELECT cid, created_at, updated_at, title, persona_id FROM webchat_conversation WHERE user_id = ? ORDER BY updated_at DESC
             ''', (user_id,)
         )
         
@@ -247,24 +278,42 @@ class SQLiteDatabase(BaseDatabase):
             cid = row[0]
             created_at = row[1]
             updated_at = row[2]
-            conversations.append(WebChatConversation("", cid, '[]', created_at, updated_at))
+            title = row[3]
+            persona_id = row[4]
+            conversations.append(Conversation("", cid, '[]', created_at, updated_at, title, persona_id))
         return conversations
     
-    def update_webchat_conversation(self, user_id: str, cid: str, history: str):
+    def update_conversation(self, user_id: str, cid: str, history: str):
+        '''更新对话，并且同时更新时间'''
+        updated_at = int(time.time())
         self._exec_sql(
             '''
-            UPDATE webchat_conversation SET history = ? WHERE user_id = ? AND cid = ?
-            ''', (history, user_id, cid)
+            UPDATE webchat_conversation SET history = ?, updated_at = ? WHERE user_id = ? AND cid = ?
+            ''', (history, updated_at, user_id, cid)
+        )
+        
+        
+    def update_conversation_title(self, user_id: str, cid: str, title: str):
+        self._exec_sql(
+            '''
+            UPDATE webchat_conversation SET title = ? WHERE user_id = ? AND cid = ?
+            ''', (title, user_id, cid)
+        )
+        
+    def update_conversation_persona_id(self, user_id: str, cid: str, persona_id: str):
+        self._exec_sql(
+            '''
+            UPDATE webchat_conversation SET persona_id = ? WHERE user_id = ? AND cid = ?
+            ''', (persona_id, user_id, cid)
         )
     
-    def delete_webchat_conversation(self, user_id: str, cid: str):
+    def delete_conversation(self, user_id: str, cid: str):
         self._exec_sql(
             '''
             DELETE FROM webchat_conversation WHERE user_id = ? AND cid = ?
             ''', (user_id, cid)
         )
 
-
     def insert_atri_vision_data(self, vision: ATRIVision):
         ts = int(time.time())
         keywords = ",".join(vision.keywords)

astrbot/core/db/sqlite_init.sql

@@ -42,5 +42,7 @@ CREATE TABLE IF NOT EXISTS webchat_conversation(
     cid TEXT,
     history TEXT,
     created_at INTEGER,
-    updated_at INTEGER
+    updated_at INTEGER,
+    title TEXT,
+    persona_id TEXT
 );

astrbot/core/pipeline/process_stage/method/llm_request.py

@@ -2,6 +2,7 @@
 本地 Agent 模式的 LLM 调用 Stage
 '''
 import traceback
+import json
 from typing import Union, AsyncGenerator
 from ...context import PipelineContext
 from ..stage import Stage
@@ -10,7 +11,7 @@ from astrbot.core.message.message_event_result import MessageEventResult, Result
 from astrbot.core.message.components import Image
 from astrbot.core import logger
 from astrbot.core.utils.metrics import Metric
-from astrbot.core.provider.entites import ProviderRequest
+from astrbot.core.provider.entites import ProviderRequest, LLMResponse
 from astrbot.core.star.star_handler import star_handlers_registry, EventType
 
 class LLMRequestSubStage(Stage):
@@ -24,6 +25,8 @@ class LLMRequestSubStage(Stage):
             if self.provider_wake_prefix.startswith(bwp):
                 logger.info(f"识别 LLM 聊天额外唤醒前缀 {self.provider_wake_prefix} 以机器人唤醒前缀 {bwp} 开头，已自动去除。")
                 self.provider_wake_prefix = self.provider_wake_prefix[len(bwp):]
+                
+        self.conv_manager = ctx.plugin_manager.context.conversation_manager
         
     async def process(self, event: AstrMessageEvent, _nested: bool = False) -> Union[None, AsyncGenerator[None, None]]:
         req: ProviderRequest = None
@@ -46,10 +49,17 @@ class LLMRequestSubStage(Stage):
                 if isinstance(comp, Image):
                     image_url = comp.url if comp.url else comp.file
                     req.image_urls.append(image_url)
-            req.session_id = event.session_id
+            
+            # 获取对话上下文
+            conversation_id = await self.conv_manager.get_curr_conversation_id(event.unified_msg_origin)
+            if not conversation_id:
+                conversation_id = await self.conv_manager.new_conversation(event.unified_msg_origin)
+            req.session_id = conversation_id
+            conversation = await self.conv_manager.get_conversation(event.unified_msg_origin, conversation_id)
+            req.conversation = conversation
+            req.contexts = json.loads(conversation.history)
+
             event.set_extra("provider_request", req)
-            session_provider_context = provider.session_memory.get(event.session_id)
-            req.contexts = session_provider_context if session_provider_context else []
             
         if not req.prompt and not req.image_urls:
             return
@@ -62,6 +72,9 @@ class LLMRequestSubStage(Stage):
                 await handler.handler(event, req)
             except BaseException:
                 logger.error(traceback.format_exc())
+                
+        if isinstance(req.contexts, str):
+            req.contexts = json.loads(req.contexts)
         
         try:
             logger.debug(f"提供商请求 Payload: {req.__dict__}")
@@ -77,6 +90,9 @@ class LLMRequestSubStage(Stage):
                 except BaseException:
                     logger.error(traceback.format_exc())
             
+            # 保存到历史记录
+            await self._save_to_history(event, req, llm_response)
+            
             await Metric.upload(llm_tick=1, model_name=provider.get_model(), provider_type=provider.meta().type)
 
             if llm_response.role == 'assistant':
@@ -117,4 +133,24 @@ class LLMRequestSubStage(Stage):
         except BaseException as e:
             logger.error(traceback.format_exc())
             event.set_result(MessageEventResult().message(f"AstrBot 请求失败。\n错误类型: {type(e).__name__}\n错误信息: {str(e)}"))
-            return
+            return
+        
+    async def _save_to_history(self, event: AstrMessageEvent, req: ProviderRequest, llm_response: LLMResponse):
+        if llm_response.role == "assistant":
+            # 文本回复
+            contexts = req.contexts
+            new_record = {
+                "role": "user",
+                "content": req.prompt
+            }
+            contexts.append(new_record)
+            contexts.append({
+                "role": "assistant",
+                "content": llm_response.completion_text
+            })
+            contexts_to_save = list(filter(lambda item: '_no_save' not in item, contexts))
+            await self.conv_manager.update_conversation(
+                event.unified_msg_origin, 
+                req.session_id, 
+                history=contexts_to_save
+            )

astrbot/core/provider/entites.py

@@ -3,6 +3,7 @@ from dataclasses import dataclass, field
 from typing import List, Dict, Type
 from .func_tool_manager import FuncCall
 from openai.types.chat.chat_completion import ChatCompletion
+from astrbot.core.db.po import Conversation
 
 
 class ProviderType(enum.Enum):
@@ -38,9 +39,9 @@ class ProviderRequest():
     '''上下文。格式与 openai 的上下文格式一致：
     参考 https://platform.openai.com/docs/api-reference/chat/create#chat-create-messages
     '''
-    
     system_prompt: str = ""
     '''系统提示词'''
+    conversation: Conversation = None
     
     
 @dataclass

astrbot/core/provider/manager.py

@@ -66,7 +66,14 @@ class ProviderManager():
         if not self.selected_default_persona and len(self.personas) > 0:
             # 默认选择第一个
             self.selected_default_persona = self.personas[0]
-        
+            
+        if not self.selected_default_persona:
+            self.selected_default_persona = Personality(
+                prompt="You are a helpful and friendly assistant.",
+                name="default",
+            )
+            self.personas.append(self.selected_default_persona)
+                
         
         self.provider_insts: List[Provider] = []
         '''加载的 Provider 的实例'''

astrbot/core/provider/provider.py

@@ -8,6 +8,8 @@ from typing import TypedDict
 from astrbot.core.provider.func_tool_manager import FuncCall
 from astrbot.core.provider.entites import LLMResponse
 from dataclasses import dataclass
+
+
 class Personality(TypedDict):
     prompt: str = ""
     name: str = ""
@@ -60,25 +62,11 @@ class Provider(AbstractProvider):
     ) -> None:
         super().__init__(provider_config)
         
-        self.session_memory = defaultdict(list)
-        '''维护了 session_id 的上下文，**不包含 system 指令**。'''
-        
         self.provider_settings = provider_settings
         
         self.curr_personality: Personality = default_persona
         '''维护了当前的使用的 persona，即人格。可能为 None'''
-        
-        self.db_helper = db_helper
-        '''用于持久化的数据库操作对象。'''
 
-        if persistant_history:
-            # 读取历史记录
-            try:
-                for history in db_helper.get_llm_history(provider_type=provider_config['id']):
-                    self.session_memory[history.session_id] = json.loads(history.content)
-            except BaseException as e:
-                logger.warning(f"读取 LLM 对话历史记录 失败：{e}。仍可正常使用。")
-    
     @abc.abstractmethod
     def get_current_key(self) -> str:
         raise NotImplementedError()
@@ -96,22 +84,6 @@ class Provider(AbstractProvider):
         '''获得支持的模型列表'''
         raise NotImplementedError()
     
-    @abc.abstractmethod
-    async def get_human_readable_context(self, session_id: str, page: int, page_size: int):
-        '''获取人类可读的上下文
-        
-        page 从 1 开始
-
-        Example:
-            
-            ["User: 你好", "Assistant: 你好！"]
-            
-        Return:
-            contexts: List[str]: 上下文列表
-            total_pages: int: 总页数
-        '''
-        raise NotImplementedError()
-    
     @abc.abstractmethod
     async def text_chat(self,
                         prompt: str,
@@ -125,26 +97,35 @@ class Provider(AbstractProvider):
         
         Args:
             prompt: 提示词
-            session_id: 会话 ID
+            session_id: 会话 ID(此属性已经被废弃)
             image_urls: 图片 URL 列表
             tools: Function-calling 工具
             contexts: 上下文
             kwargs: 其他参数
             
         Notes:
-            - 如果传入了 contexts，将会提前加上上下文。否则使用 session_memory 中的上下文。
-            - 可以选择性地传入 session_id，如果传入了 session_id，将会使用 session_id 对应的上下文进行对话，
-            并且也会记录相应的对话上下文，实现多轮对话。如果不传入则不会记录上下文。
             - 如果传入了 image_urls，将会在对话时附上图片。如果模型不支持图片输入，将会抛出错误。
             - 如果传入了 tools，将会使用 tools 进行 Function-calling。如果模型不支持 Function-calling，将会抛出错误。
         '''
         raise NotImplementedError()
-    
-    @abc.abstractmethod
-    async def forget(self, session_id: str) -> bool:
-        '''重置某一个 session_id 的上下文'''
-        raise NotImplementedError()
 
+    async def pop_record(self, context: List):
+        '''
+        弹出 context 第一条非系统提示词对话记录
+        '''
+        poped = 0
+        indexs_to_pop = []
+        for idx, record in enumerate(context):
+            if record["role"] == "system":
+                continue
+            else:
+                indexs_to_pop.append(idx)
+                poped += 1
+                if poped == 2:
+                    break
+                
+        for idx in reversed(indexs_to_pop):
+            context.pop(idx)
 
         
 class STTProvider(AbstractProvider):

astrbot/core/provider/sources/gemini_source.py

@@ -73,60 +73,10 @@ class ProviderGoogleGenAI(Provider):
             api_base=provider_config.get("api_base", None)
         )
         self.set_model(provider_config['model_config']['model'])
-    
-    async def get_human_readable_context(self, session_id, page, page_size):
-        if session_id not in self.session_memory:
-            raise Exception("会话 ID 不存在")
-        contexts = []
-        temp_contexts = []
-        for record in self.session_memory[session_id]:
-            if record['role'] == "user":
-                temp_contexts.append(f"User: {record['content']}")
-            elif record['role'] == "assistant":
-                temp_contexts.append(f"Assistant: {record['content']}")
-                contexts.insert(0, temp_contexts)
-                temp_contexts = []
-
-        # 展平 contexts 列表
-        contexts = [item for sublist in contexts for item in sublist]
-
-        # 计算分页
-        paged_contexts = contexts[(page-1)*page_size:page*page_size]
-        total_pages = len(contexts) // page_size
-        if len(contexts) % page_size != 0:
-            total_pages += 1
-        
-        return paged_contexts, total_pages
 
     async def get_models(self):
         return await self.client.models_list()
     
-    async def pop_record(self, session_id: str, pop_system_prompt: bool = False):
-        '''
-        弹出第一条记录
-        '''
-        if session_id not in self.session_memory:
-            raise Exception("会话 ID 不存在")
-
-        if len(self.session_memory[session_id]) == 0:
-            return None
-
-        for i in range(len(self.session_memory[session_id])):
-            # 检查是否是 system prompt
-            if not pop_system_prompt and self.session_memory[session_id][i]['user']['role'] == "system":
-                # 如果只有一个 system prompt，才不删掉
-                f = False
-                for j in range(i+1, len(self.session_memory[session_id])):
-                    if self.session_memory[session_id][j]['user']['role'] == "system":
-                        f = True
-                        break
-                if not f:
-                    continue
-            record = self.session_memory[session_id].pop(i)
-            break
-
-        return record
-    
     async def _query(self, payloads: dict, tools: FuncCall) -> LLMResponse:
         tool = None
         if tools:
@@ -197,11 +147,10 @@ class ProviderGoogleGenAI(Provider):
         llm_response.completion_text = llm_response.completion_text.strip()
         return llm_response
 
-
     async def text_chat(
         self,
         prompt: str,
-        session_id: str,
+        session_id: str = None,
         image_urls: List[str]=None,
         func_tool: FuncCall=None,
         contexts=None,
@@ -210,10 +159,7 @@ class ProviderGoogleGenAI(Provider):
     ) -> LLMResponse: 
         new_record = await self.assemble_context(prompt, image_urls)
         context_query = []
-        if not contexts:
-            context_query = [*self.session_memory[session_id], new_record]
-        else:
-            context_query = [*contexts, new_record]
+        context_query = [*contexts, new_record]
         if system_prompt:
             context_query.insert(0, {"role": "system", "content": system_prompt})
             
@@ -234,7 +180,7 @@ class ProviderGoogleGenAI(Provider):
                 while retry_cnt > 0:
                     logger.warning(f"请求失败：{e}。上下文长度超过限制。尝试弹出最早的记录然后重试。")
                     try:
-                        self.pop_record(session_id)
+                        await self.pop_record(context_query)
                         llm_response = await self._query(payloads, func_tool)
                         break
                     except Exception as e:
@@ -254,34 +200,7 @@ class ProviderGoogleGenAI(Provider):
                 
                 raise e
 
-        if kwargs.get("persist", True) and llm_response:
-            await self.save_history(contexts, new_record, session_id, llm_response)
-            
         return llm_response
-    
-    async def save_history(self, contexts: List, new_record: dict, session_id: str, llm_response: LLMResponse):
-        if llm_response.role == "assistant" and session_id:
-            # 文本回复
-            if not contexts:
-                # 添加用户 record
-                self.session_memory[session_id].append(new_record)
-                # 添加 assistant record
-                self.session_memory[session_id].append({
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                })
-            else:
-                contexts_to_save = list(filter(lambda item: '_no_save' not in item, contexts))
-                self.session_memory[session_id] = [*contexts_to_save, new_record, {
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        
-    async def forget(self, session_id: str) -> bool:
-        self.session_memory[session_id] = []
-        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        return True
 
     def get_current_key(self) -> str:
         return self.client.api_key

astrbot/core/provider/sources/llmtuner_source.py

@@ -63,14 +63,7 @@ class LLMTunerModelLoader(Provider):
     ) -> LLMResponse:
         system_prompt = ""
         new_record = {"role": "user", "content": prompt}
-        if not contexts:
-            query_context = [
-                *self.session_memory[session_id],
-                new_record,
-            ]
-            system_prompt = self.curr_personality["prompt"]
-        else:
-            query_context = [*contexts, new_record]
+        query_context = [*contexts, new_record]
 
         # 提取出系统提示
         system_idxs = []
@@ -96,34 +89,8 @@ class LLMTunerModelLoader(Provider):
         responses = await self.model.achat(**conf)
 
         llm_response = LLMResponse("assistant", responses[-1].response_text)
-        
-        await self.save_history(contexts, new_record, session_id, llm_response)
-        
+                
         return llm_response
-   
-    async def save_history(self, contexts: List, new_record: dict, session_id: str, llm_response: LLMResponse):
-        if llm_response.role == "assistant" and session_id:
-            # 文本回复
-            if not contexts:
-                # 添加用户 record
-                self.session_memory[session_id].append(new_record)
-                # 添加 assistant record
-                self.session_memory[session_id].append({
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                })
-            else:
-                contexts_to_save = list(filter(lambda item: '_no_save' not in item, contexts))
-                self.session_memory[session_id] = [*contexts_to_save, new_record, {
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        
-    async def forget(self, session_id):
-        self.session_memory[session_id] = []
-        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        return True
 
     async def get_current_key(self):
         return "none"
@@ -132,28 +99,4 @@ class LLMTunerModelLoader(Provider):
         pass
 
     async def get_models(self):
-        return [self.get_model()]
-
-    async def get_human_readable_context(self, session_id, page, page_size):
-        if session_id not in self.session_memory:
-            raise Exception("会话 ID 不存在")
-        contexts = []
-        temp_contexts = []
-        for record in self.session_memory[session_id]:
-            if record["role"] == "user":
-                temp_contexts.append(f"User: {record['content']}")
-            elif record["role"] == "assistant":
-                temp_contexts.append(f"Assistant: {record['content']}")
-                contexts.insert(0, temp_contexts)
-                temp_contexts = []
-
-        # 展平 contexts 列表
-        contexts = [item for sublist in contexts for item in sublist]
-
-        # 计算分页
-        paged_contexts = contexts[(page - 1) * page_size : page * page_size]
-        total_pages = len(contexts) // page_size
-        if len(contexts) % page_size != 0:
-            total_pages += 1
-
-        return paged_contexts, total_pages
+        return [self.get_model()]

astrbot/core/provider/sources/openai_source.py

@@ -48,30 +48,6 @@ class ProviderOpenAIOfficial(Provider):
             )
             
         self.set_model(provider_config['model_config']['model'])
-    
-    async def get_human_readable_context(self, session_id, page, page_size):
-        if session_id not in self.session_memory:
-            raise Exception("会话 ID 不存在")
-        contexts = []
-        temp_contexts = []
-        for record in self.session_memory[session_id]:
-            if record['role'] == "user":
-                temp_contexts.append(f"User: {record['content']}")
-            elif record['role'] == "assistant":
-                temp_contexts.append(f"Assistant: {record['content']}")
-                contexts.insert(0, temp_contexts)
-                temp_contexts = []
-
-        # 展平 contexts 列表
-        contexts = [item for sublist in contexts for item in sublist]
-
-        # 计算分页
-        paged_contexts = contexts[(page-1)*page_size:page*page_size]
-        total_pages = len(contexts) // page_size
-        if len(contexts) % page_size != 0:
-            total_pages += 1
-        
-        return paged_contexts, total_pages
 
     async def get_models(self):
         try:
@@ -84,22 +60,6 @@ class ProviderOpenAIOfficial(Provider):
         except NotFoundError as e:
             raise Exception(f"获取模型列表失败：{e}")
     
-    async def pop_record(self, session_id: str):
-        '''
-        弹出最早的一个对话
-        '''
-        if session_id not in self.session_memory:
-            raise Exception("会话 ID 不存在")
-
-        if len(self.session_memory[session_id]) < 2:
-            return
-        
-        try:
-            self.session_memory[session_id].pop(0)
-            self.session_memory[session_id].pop(0)
-        except IndexError:
-            pass
-    
     async def _query(self, payloads: dict, tools: FuncCall) -> LLMResponse:
         if tools:
             tool_list = tools.get_func_desc_openai_style()
@@ -141,7 +101,7 @@ class ProviderOpenAIOfficial(Provider):
     async def text_chat(
         self,
         prompt: str,
-        session_id: str,
+        session_id: str=None,
         image_urls: List[str]=None,
         func_tool: FuncCall=None,
         contexts=None,
@@ -149,11 +109,7 @@ class ProviderOpenAIOfficial(Provider):
         **kwargs
     ) -> LLMResponse: 
         new_record = await self.assemble_context(prompt, image_urls)
-        context_query = []
-        if not contexts:
-            context_query = [*self.session_memory[session_id], new_record]
-        else:
-            context_query = [*contexts, new_record]
+        context_query = [*contexts, new_record]
         if system_prompt:
             context_query.insert(0, {"role": "system", "content": system_prompt})
 
@@ -214,9 +170,6 @@ class ProviderOpenAIOfficial(Provider):
                 
                 raise e
         
-        if kwargs.get("persist", True) and llm_response:
-            await self.save_history(contexts, new_record, session_id, llm_response)
-        
         return llm_response
     
     async def _remove_image_from_context(self, contexts: List):
@@ -244,32 +197,7 @@ class ProviderOpenAIOfficial(Provider):
                 context['content'] = new_content
             new_contexts.append(context)
         return new_contexts
-
     
-    async def save_history(self, contexts: List, new_record: dict, session_id: str, llm_response: LLMResponse):
-        if llm_response.role == "assistant" and session_id:
-            # 文本回复
-            if not contexts:
-                # 添加用户 record
-                self.session_memory[session_id].append(new_record)
-                # 添加 assistant record
-                self.session_memory[session_id].append({
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                })
-            else:
-                contexts_to_save = list(filter(lambda item: '_no_save' not in item, contexts))
-                self.session_memory[session_id] = [*contexts_to_save, new_record, {
-                    "role": "assistant",
-                    "content": llm_response.completion_text
-                }]
-            self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        
-    async def forget(self, session_id: str) -> bool:
-        self.session_memory[session_id] = []
-        self.db_helper.update_llm_history(session_id, json.dumps(self.session_memory[session_id]), self.provider_config['id'])
-        return True
-
     def get_current_key(self) -> str:
         return self.client.api_key
 

astrbot/core/provider/sources/zhipu_source.py

@@ -22,7 +22,7 @@ class ProviderZhipu(ProviderOpenAIOfficial):
     async def text_chat(
         self,
         prompt: str,
-        session_id: str,
+        session_id: str = None,
         image_urls: List[str]=None,
         func_tool: FuncCall=None,
         contexts=None,
@@ -32,10 +32,7 @@ class ProviderZhipu(ProviderOpenAIOfficial):
         new_record = await self.assemble_context(prompt, image_urls)
         context_query = []
         
-        if not contexts:
-            context_query = [*self.session_memory[session_id], new_record]
-        else:
-            context_query = [*contexts, new_record]
+        context_query = [*contexts, new_record]
         
         model_cfgs: dict = self.provider_config.get("model_config", {})
         # glm-4v-flash 只支持一张图片
@@ -62,7 +59,6 @@ class ProviderZhipu(ProviderOpenAIOfficial):
         }
         try:
             llm_response = await self._query(payloads, func_tool)
-            await self.save_history(contexts, new_record, session_id, llm_response)
             return llm_response
         except Exception as e:
             if "maximum context length" in str(e):

astrbot/core/star/context.py

@@ -16,6 +16,7 @@ from .filter.command import CommandFilter
 from .filter.regex import RegexFilter
 from typing import Awaitable
 from astrbot.core.rag.knowledge_db_mgr import KnowledgeDBManager
+from astrbot.core.conversation_mgr import ConversationManager
 
 class Context:
     '''
@@ -44,6 +45,7 @@ class Context:
         db: BaseDatabase, 
         provider_manager: ProviderManager = None, 
         platform_manager: PlatformManager = None,
+        conversation_manager: ConversationManager = None,
         knowledge_db_manager: KnowledgeDBManager = None
     ):
         self._event_queue = event_queue
@@ -52,6 +54,7 @@ class Context:
         self.provider_manager = provider_manager
         self.platform_manager = platform_manager
         self.knowledge_db_manager = knowledge_db_manager
+        self.conversation_manager = conversation_manager
 
     def get_registered_star(self, star_name: str) -> StarMetadata:
         '''根据插件名获取插件的 Metadata'''

astrbot/core/utils/param_validation_mixin.py

@@ -25,6 +25,10 @@ class ParameterValidationMixin:
                     elif isinstance(param_type_or_default_val, str):
                         # 如果 param_type_or_default_val 是字符串，直接赋值
                         result[param_name] = params[i]
+                    elif isinstance(param_type_or_default_val, int):
+                        result[param_name] = int(params[i])
+                    elif isinstance(param_type_or_default_val, float):
+                        result[param_name] = float(params[i])
                     else:
                         result[param_name] = param_type_or_default_val(params[i])
                 except ValueError:

astrbot/dashboard/routes/chat.py

@@ -121,7 +121,7 @@ class ChatRoute(Route):
         }))
         
         # 持久化
-        conversation = self.db.get_webchat_conversation_by_user_id(username, conversation_id)
+        conversation = self.db.get_conversation_by_user_id(username, conversation_id)
         try:
             history = json.loads(conversation.history)
         except BaseException as e:
@@ -136,7 +136,7 @@ class ChatRoute(Route):
         if audio_url:
             new_his['audio_url'] = audio_url
         history.append(new_his)
-        self.db.update_webchat_conversation(username, conversation_id, history=json.dumps(history))
+        self.db.update_conversation(username, conversation_id, history=json.dumps(history))
         
         return Response().ok().__dict__
         
@@ -168,7 +168,7 @@ class ChatRoute(Route):
                         continue
                     yield result_text + '\n'
                     
-                    conversation = self.db.get_webchat_conversation_by_user_id(username, cid)
+                    conversation = self.db.get_conversation_by_user_id(username, cid)
                     try:
                         history = json.loads(conversation.history)
                     except BaseException as e:
@@ -178,7 +178,7 @@ class ChatRoute(Route):
                         'type': 'bot',
                         'message': result_text
                     })
-                    self.db.update_webchat_conversation(username, cid, history=json.dumps(history))
+                    self.db.update_conversation(username, cid, history=json.dumps(history))
                     
                     await asyncio.sleep(0.5)
             except BaseException as e:
@@ -204,20 +204,20 @@ class ChatRoute(Route):
         if not conversation_id:
             return Response().error("Missing key: conversation_id").__dict__
         
-        self.db.delete_webchat_conversation(username, conversation_id)
+        self.db.delete_conversation(username, conversation_id)
         return Response().ok().__dict__
     
     async def new_conversation(self):
         username = g.get('username', 'guest')
         conversation_id = str(uuid.uuid4())
-        self.db.webchat_new_conversation(username, conversation_id)
+        self.db.new_conversation(username, conversation_id)
         return Response().ok(data={
             'conversation_id': conversation_id
         }).__dict__
         
     async def get_conversations(self):
         username = g.get('username', 'guest')
-        conversations = self.db.get_webchat_conversations(username)
+        conversations = self.db.get_conversations(username)
         return Response().ok(data=conversations).__dict__
     
     async def get_conversation(self):
@@ -226,7 +226,7 @@ class ChatRoute(Route):
         if not conversation_id:
             return Response().error("Missing key: conversation_id").__dict__
         
-        conversation = self.db.get_webchat_conversation_by_user_id(username, conversation_id)
+        conversation = self.db.get_conversation_by_user_id(username, conversation_id)
         
         self.curr_user_cid[username] = conversation_id
         

packages/astrbot/main.py

@@ -1,6 +1,7 @@
 import aiohttp
 import datetime
 import builtins
+import json
 import astrbot.api.star as star
 import astrbot.api.event.filter as filter
 from astrbot.api.event import AstrMessageEvent, MessageEventResult
@@ -59,24 +60,28 @@ AstrBot 指令:
 /plugin: 查看插件、插件帮助
 /t2i: 开关文本转图片
 /sid: 获取会话 ID
-/op <admin_id>: 授权管理员
-/deop <admin_id>: 取消管理员
-/wl <sid>: 添加白名单
-/dwl <sid>: 删除白名单
-/dashboard_update: 更新管理面板
-/alter_cmd: 设置指令权限
+/op <admin_id>: 授权管理员(op)
+/deop <admin_id>: 取消管理员(op)
+/wl <sid>: 添加白名单(op)
+/dwl <sid>: 删除白名单(op)
+/dashboard_update: 更新管理面板(op)
+/alter_cmd: 设置指令权限(op)
 
 [大模型]
 /provider: 大模型提供商
 /model: 模型列表
-/key: API Key
-/reset: 重置 LLM 会话
-/history: 对话记录
-/persona: 人格情景
+/ls: 对话列表
+/new: 创建新对话
+/switch: 切换对话
+/del: 删除当前会话对话(op)
+/reset: 重置 LLM 会话(op)
+/history: 当前对话的对话记录
+/persona: 人格情景(op)
 /tool ls: 函数工具
+/key: API Key(op)
 
 [其他]
-/set <变量名> <值>: 为会话定义一个变量。适用于 Dify 工作流输入。
+/set <变量名> <值>: 为会话定义变量。适用于 Dify 工作流输入。
 /unset <变量名>: 删除会话的变量。
 
 提示：如要查看插件指令，请输入 /plugin 查看具体信息。
@@ -273,7 +278,10 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
             message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
             return
         
-        await self.context.get_using_provider().forget(message.session_id)
+        await self.context.conversation_manager.update_conversation(
+            message.unified_msg_origin, message.session_id, []
+        )
+        
         ret = "清除会话 LLM 聊天历史成功。"
         if self.ltm:
             cnt = await self.ltm.remove_session(event=message)
@@ -329,20 +337,22 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
 
     @filter.command("history")
     async def his(self, message: AstrMessageEvent, page: int = 1):
-        
-                
+        '''查看对话记录'''
         if not self.context.get_using_provider():
             message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
             return
         
-        size_per_page = 3
-        contexts, total_pages = await self.context.get_using_provider().get_human_readable_context(message.session_id, page, size_per_page)
+        size_per_page = 6
+        session_curr_cid = await self.context.conversation_manager.get_curr_conversation_id(message.unified_msg_origin)
+        contexts, total_pages = await self.context.conversation_manager.get_human_readable_context(
+            message.unified_msg_origin, session_curr_cid, page, size_per_page
+        )
 
         history = ""
         for context in contexts:
             history += f"{context}\n"
             
-        ret = f"""历史记录：
+        ret = f"""当前对话历史记录：
 {history}
 第 {page} 页 | 共 {total_pages} 页
 
@@ -351,6 +361,64 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
 
         message.set_result(MessageEventResult().message(ret).use_t2i(False))
 
+    @filter.command("ls")
+    async def convs(self, message: AstrMessageEvent, page: int = 1):
+        '''查看对话列表'''
+        size_per_page = 6
+        conversations = await self.context.conversation_manager.get_conversations(message.unified_msg_origin)
+        total_pages = len(conversations) // size_per_page
+        if len(conversations) % size_per_page != 0:
+            total_pages += 1
+        conversations = conversations[(page-1)*size_per_page:page*size_per_page]
+        
+        ret = "\n对话列表：\n"
+        global_index = (page - 1) * size_per_page + 1
+        
+        for conv in conversations:
+            
+            persona_id = conv.persona_id
+            if not persona_id and not persona_id == "[%None]":
+                persona_id = self.context.provider_manager.selected_default_persona['name']
+            
+            ret += f"{global_index}. 新对话{conv.cid[:4]}\n  人格情景: {persona_id}\n上次更新: {datetime.datetime.fromtimestamp(conv.updated_at).strftime('%m-%d %H:%M')}\n"
+            global_index += 1
+        
+        curr_cid = await self.context.conversation_manager.get_curr_conversation_id(message.unified_msg_origin)
+        if curr_cid:
+            ret += f"\n当前对话: {curr_cid[:4]}"
+        else:
+            ret += "\n当前对话: 无"
+        ret += f"\n第 {page} 页 | 共 {total_pages} 页"
+        ret += "\n*输入 /ls 2 跳转到第 2 页"
+        
+        message.set_result(MessageEventResult().message(ret).use_t2i(False))
+        
+    @filter.command("new")
+    async def new_conv(self, message: AstrMessageEvent):
+        '''创建新对话'''
+        cid = await self.context.conversation_manager.new_conversation(message.unified_msg_origin)
+        message.set_result(MessageEventResult().message(f"切换到新对话: {cid[:4]}。"))
+            
+    @filter.command("switch")
+    async def switch_conv(self, message: AstrMessageEvent, index: int):
+        '''切换对话'''
+        conversations = await self.context.conversation_manager.get_conversations(message.unified_msg_origin)
+        if index > len(conversations) or index < 1:
+            message.set_result(MessageEventResult().message("对话序号错误。"))
+        else:
+            conversation = conversations[index-1]
+            await self.context.conversation_manager.switch_conversation(message.unified_msg_origin, conversation.cid)
+            message.set_result(MessageEventResult().message(f"切换到对话: {conversation.cid[:4]}。"))
+    
+    @filter.permission_type(filter.PermissionType.ADMIN)
+    @filter.command("del")
+    async def del_conv(self, message: AstrMessageEvent):
+        '''删除当前对话'''
+        session_curr_cid = await self.context.conversation_manager.get_curr_conversation_id(message.unified_msg_origin)
+        await self.context.conversation_manager.delete_conversation(message.unified_msg_origin, session_curr_cid)
+        message.set_result(MessageEventResult().message("删除当前对话成功。"))
+        
+
     @filter.permission_type(filter.PermissionType.ADMIN)
     @filter.command("key")
     async def key(self, message: AstrMessageEvent, index: int=None):
@@ -387,28 +455,28 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
     @filter.permission_type(filter.PermissionType.ADMIN)
     @filter.command("persona")
     async def persona(self, message: AstrMessageEvent):
-                
-        if not self.context.get_using_provider():
-            message.set_result(MessageEventResult().message("未找到任何 LLM 提供商。请先配置。"))
-            return
-        
-        
         l = message.message_str.split(" ")
         
         curr_persona_name = "无"
-        if self.context.get_using_provider().curr_personality:
-            curr_persona_name = self.context.get_using_provider().curr_personality['name']
+        cid = await self.context.conversation_manager.get_curr_conversation_id(message.unified_msg_origin)
+        if cid:
+            conversation = await self.context.conversation_manager.get_conversation(message.unified_msg_origin, cid)
+            if not conversation.persona_id and not conversation.persona_id == "[%None]":
+                curr_persona_name = self.context.provider_manager.selected_default_persona['name']
+            else:
+                curr_persona_name = conversation.persona_id
         
         if len(l) == 1:
             message.set_result(
                 MessageEventResult().message(f"""[Persona]
 
-- 设置人格情景: `/persona 人格名`, 如 /persona 编剧
 - 人格情景列表: `/persona list`
-- 人格情景详细信息: `/persona view 人格名`
+- 设置人格情景: `/persona 人格`
+- 人格情景详细信息: `/persona view 人格`
 - 取消人格: `/persona unset`
 
-当前人格情景: {curr_persona_name}
+默认人格情景: {self.context.provider_manager.selected_default_persona['name']}
+当前对话 {cid[:4]} 的人格情景: {curr_persona_name}
 
 配置人格情景请前往管理面板-配置页
 """).use_t2i(False))
@@ -433,7 +501,10 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                 msg = f"人格{ps}不存在"
             message.set_result(MessageEventResult().message(msg))
         elif l[1] == "unset":
-            self.context.get_using_provider().curr_personality = None
+            if not cid:
+                message.set_result(MessageEventResult().message("当前没有对话，无法取消人格。"))
+                return
+            await self.context.conversation_manager.update_conversation_persona_id(message.unified_msg_origin, "[%None]")
             message.set_result(MessageEventResult().message("取消人格成功。"))
         else:
             ps = "".join(l[1:]).strip()
@@ -441,7 +512,7 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                 lambda persona: persona['name'] == ps, 
                 self.context.provider_manager.personas
             ), None):
-                self.context.get_using_provider().curr_personality = persona
+                await self.context.conversation_manager.update_conversation_persona_id(message.unified_msg_origin, ps)
                 message.set_result(MessageEventResult().message("设置成功。如果您正在切换到不同的人格，请注意使用 /reset 来清空上下文，防止原人格对话影响现人格。"))
             else:
                 message.set_result(MessageEventResult().message("不存在该人格情景。使用 /persona list 查看所有。"))
@@ -513,7 +584,12 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                     logger.error("未找到任何 LLM 提供商。请先配置。无法主动回复")
                     return
                 try:
-                    session_provider_context = provider.session_memory.get(event.session_id)
+                    session_curr_cid = await self.context.conversation_manager.get_curr_conversation_id(event.unified_msg_origin)
+                    conv = await self.context.conversation_manager.get_conversation(
+                        event.unified_msg_origin, 
+                        session_curr_cid
+                    )
+                    history = json.loads(conv.history)
                     
                     prompt = self.ltm.ar_prompt
                     if not prompt:
@@ -523,7 +599,7 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
                         prompt=prompt,
                         func_tool_manager=self.context.get_llm_tool_manager(),
                         session_id=event.session_id,
-                        contexts=session_provider_context if session_provider_context else []
+                        contexts=history if history else []
                     )
                 except BaseException as e:
                     logger.error(f"主动回复失败: {e}")
@@ -545,14 +621,22 @@ UID: {user_id} 此 ID 可用于设置管理员。/op <UID> 授权管理员, /deo
         if self.enable_datetime:
             req.system_prompt += f"\nCurrent datetime: {datetime.datetime.now().strftime('%Y-%m-%d %H:%M')}\n"
         
-        if persona := provider.curr_personality:
-            if prompt := persona['prompt']:
-                req.system_prompt += prompt
-            if mood_dialogs := persona['_mood_imitation_dialogs_processed']:
-                req.system_prompt += "\nHere are few shots of dialogs, you need to imitate the tone of 'B' in the following dialogs to respond:\n"
-                req.system_prompt += mood_dialogs
-            if begin_dialogs := persona["_begin_dialogs_processed"]:
-                req.contexts[:0] = begin_dialogs
+        if req.conversation:
+            persona_id = req.conversation.persona_id
+            if not persona_id and persona_id != "[%None]": # [%None] 为用户取消人格
+                persona_id = self.context.provider_manager.selected_default_persona['name']
+            persona = next(builtins.filter(
+                lambda persona: persona['name'] == persona_id, 
+                self.context.provider_manager.personas
+            ), None)
+            if persona:
+                if prompt := persona['prompt']:
+                    req.system_prompt += prompt
+                if mood_dialogs := persona['_mood_imitation_dialogs_processed']:
+                    req.system_prompt += "\nHere are few shots of dialogs, you need to imitate the tone of 'B' in the following dialogs to respond:\n"
+                    req.system_prompt += mood_dialogs
+                if begin_dialogs := persona["_begin_dialogs_processed"]:
+                    req.contexts[:0] = begin_dialogs
                 
         if self.ltm:
             try: