feat: 增加模型响应后的插件钩子

remove: 移除了默认的r1过滤

astrbot/api/event/filter/__init__.py

@@ -6,6 +6,7 @@ from astrbot.core.star.register import (
     register_platform_adapter_type as platform_adapter_type,
     register_permission_type as permission_type,
     register_on_llm_request as on_llm_request,
+    register_on_llm_response as on_llm_response,
     register_llm_tool as llm_tool,
     register_on_decorating_result as on_decorating_result,
     register_after_message_sent as after_message_sent
@@ -31,5 +32,6 @@ __all__ = [
     'on_llm_request',
     'llm_tool',
     'on_decorating_result',
-    'after_message_sent'
+    'after_message_sent',
+    'on_llm_response'
 ]

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

@@ -68,6 +68,15 @@ class LLMRequestSubStage(Stage):
             if _nested:
                 req.func_tool = None # 暂时不支持递归工具调用
             llm_response = await provider.text_chat(**req.__dict__) # 请求 LLM
+            
+            # 执行 LLM 响应后的事件。
+            handlers = star_handlers_registry.get_handlers_by_event_type(EventType.OnLLMResponseEvent)
+            for handler in handlers:
+                try:
+                    await handler.handler(event, llm_response)
+                except BaseException:
+                    logger.error(traceback.format_exc())
+            
             await Metric.upload(llm_tick=1, model_name=provider.get_model(), provider_type=provider.meta().type)
 
             if llm_response.role == 'assistant':

astrbot/core/provider/entites.py

@@ -2,6 +2,7 @@ import enum
 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
 
 
 class ProviderType(enum.Enum):
@@ -51,4 +52,6 @@ class LLMResponse:
     tools_call_args: List[Dict[str, any]] = field(default_factory=list)
     '''工具调用参数'''
     tools_call_name: List[str] = field(default_factory=list)
-    '''工具调用名称'''
+    '''工具调用名称'''
+    
+    raw_completion: ChatCompletion = None

astrbot/core/provider/sources/openai_source.py

@@ -101,17 +101,16 @@ class ProviderOpenAIOfficial(Provider):
                 stream=False
             )
         except BaseException as e:
-            if 'does not support Function Calling' in e \
-                or 'does not support tools' in e: # ollama 
+            if 'does not support Function Calling' in str(e) \
+                or 'does not support tools' in str(e): # ollama 
                     del payloads['tools']
                     logger.debug(f"模型 {self.model_name} 不支持 tools，已自动移除")
                     completion = await self.client.chat.completions.create(
                         **payloads,
                         stream=False
                     )
-        
-        if not completion:
-            raise Exception("API 返回的 completion 为空。")
+            else:
+                raise e
         
         assert isinstance(completion, ChatCompletion)
         logger.debug(f"completion: {completion}")
@@ -123,14 +122,8 @@ class ProviderOpenAIOfficial(Provider):
         if choice.message.content:
             # text completion
             completion_text = str(choice.message.content).strip()
-            
-            # 适配 deepseek-r1 模型
-            if r'<think>' in completion_text or r'</think>' in completion_text:
-                completion_text = re.sub(r'<think>.*?</think>', '', completion_text, flags=re.DOTALL).strip()
-                # 可能有单标签情况
-                completion_text = completion_text.replace(r'<think>', '').replace(r'</think>', '').strip()
-            
-            return LLMResponse("assistant", completion_text)
+
+            return LLMResponse("assistant", completion_text, raw_completion=completion)
         elif choice.message.tool_calls:
             # tools call (function calling)
             args_ls = []
@@ -141,8 +134,9 @@ class ProviderOpenAIOfficial(Provider):
                         args = json.loads(tool_call.function.arguments)
                         args_ls.append(args)
                         func_name_ls.append(tool_call.function.name)
-            return LLMResponse(role="tool", tools_call_args=args_ls, tools_call_name=func_name_ls)
+            return LLMResponse(role="tool", tools_call_args=args_ls, tools_call_name=func_name_ls, raw_completion=completion)
         else:
+            logger.error(f"API 返回的 completion 无法解析：{completion}。")
             raise Exception("Internal Error")
 
     async def text_chat(
@@ -195,6 +189,7 @@ class ProviderOpenAIOfficial(Provider):
                         else:
                             raise e
             else:
+                logger.error(f"发生了错误。Provider 配置如下: {self.provider_config}")
                 raise e
 
     

astrbot/core/star/register/__init__.py

@@ -7,6 +7,7 @@ from .star_handler import (
     register_regex,
     register_permission_type,
     register_on_llm_request,
+    register_on_llm_response,
     register_llm_tool,
     register_on_decorating_result,
     register_after_message_sent
@@ -21,6 +22,7 @@ __all__ = [
     'register_regex',
     'register_permission_type',
     'register_on_llm_request',
+    'register_on_llm_response',
     'register_llm_tool',
     'register_on_decorating_result',
     'register_after_message_sent'

astrbot/core/star/register/star_handler.py

@@ -139,6 +139,8 @@ def register_on_llm_request():
     
     Examples:
     ```py
+    from astrbot.api.provider import ProviderRequest
+    
     @on_llm_request()
     async def test(self, event: AstrMessageEvent, request: ProviderRequest) -> None:
         request.system_prompt += "你是一个猫娘..."
@@ -152,6 +154,27 @@ def register_on_llm_request():
     
     return decorator
 
+def register_on_llm_response():
+    '''当有 LLM 请求后的事件
+    
+    Examples:
+    ```py
+    from astrbot.api.provider import LLMResponse
+    
+    @on_llm_response()
+    async def test(self, event: AstrMessageEvent, response: LLMResponse) -> None:
+        ...
+    ```
+    
+    请务必接收两个参数：event, request
+    '''
+    def decorator(awaitable):
+        _ = get_handler_or_create(awaitable, EventType.OnLLMResponseEvent)
+        return awaitable
+    
+    return decorator
+
+
 def register_llm_tool(name: str = None):
     '''为函数调用（function-calling / tools-use）添加工具。
     

astrbot/core/star/star_handler.py

@@ -47,6 +47,7 @@ class EventType(enum.Enum):
     '''
     AdapterMessageEvent = enum.auto() # 收到适配器发来的消息
     OnLLMRequestEvent = enum.auto() # 收到 LLM 请求（可以是用户也可以是插件）
+    OnLLMResponseEvent = enum.auto() # LLM 响应后
     OnDecoratingResultEvent = enum.auto() # 发送消息前
     OnCallingFuncToolEvent = enum.auto() # 调用函数工具
     OnAfterMessageSentEvent = enum.auto() # 发送消息后