Fix/fix: resolve MCP tools race condition causing 'completion 无法解析' error

astrbot/core/provider/func_tool_manager.py

@@ -4,6 +4,7 @@
 import copy
 import json
 import os
+import urllib.parse
 from collections.abc import AsyncGenerator, Awaitable, Callable
 from typing import Any
 
@@ -212,37 +213,83 @@ async def init_mcp_clients(self) -> None:
             open(mcp_json_file, encoding="utf-8"),
         )["mcpServers"]
 
-        for name in mcp_server_json_obj:
-            cfg = mcp_server_json_obj[name]
+        tasks: dict[str, asyncio.Task] = {}
+
+        for name, cfg in mcp_server_json_obj.items():
             if cfg.get("active", True):
                 event = asyncio.Event()
-                asyncio.create_task(
+                task = asyncio.create_task(
                     self._init_mcp_client_task_wrapper(name, cfg, event),
                 )
+                tasks[name] = task
                 self.mcp_client_event[name] = event
 
+        if tasks:
+            logger.info(f"等待 {len(tasks)} 个 MCP 服务初始化...")
+
+            done, pending = await asyncio.wait(tasks.values(), timeout=20.0)
+
+            if pending:
+                logger.warning(
+                    "MCP 服务初始化超时（20秒），部分服务可能未完全加载。"
+                    "建议检查 MCP 服务器配置和网络连接。"
+                )
+                for task in pending:
+                    task.cancel()
+
+            success_count = 0
+            failed_services: list[str] = []
+
+            for name, task in tasks.items():
+                if task in pending:
+                    logger.error(f"MCP 服务 {name} 初始化超时")
+                    failed_services.append(name)
+                    continue
+
+                exc = task.exception()
+                if exc is not None:
+                    logger.error(f"MCP 服务 {name} 初始化失败: {exc}")
+                    # 仅在 debug 级别输出完整配置，避免在生产日志中泄露敏感信息
+                    cfg = mcp_server_json_obj.get(name, {})
+                    if "command" in cfg:
+                        logger.debug(f"  命令: {cfg['command']}")
+                        if "args" in cfg:
+                            logger.debug(f"  参数: {cfg['args']}")
+                    elif "url" in cfg:
+                        parsed = urllib.parse.urlparse(cfg["url"])
+                        logger.debug(f"  主机: {parsed.scheme}://{parsed.netloc}")
+                    failed_services.append(name)
+                else:
+                    success_count += 1
+
+            if failed_services:
+                logger.warning(
+                    f"以下 MCP 服务初始化失败: {', '.join(failed_services)}。"
+                    f"请检查配置文件 mcp_server.json 和服务器可用性。"
+                )
+
+            logger.info(f"MCP 服务初始化完成: {success_count}/{len(tasks)} 成功")
+
     async def _init_mcp_client_task_wrapper(
         self,
         name: str,
         cfg: dict,
         event: asyncio.Event,
-        ready_future: asyncio.Future | None = None,
     ) -> None:
         """初始化 MCP 客户端的包装函数，用于捕获异常"""
+        initialized = False
         try:
             await self._init_mcp_client(name, cfg)
-            tools = await self.mcp_client_dict[name].list_tools_and_save()
-            if ready_future and not ready_future.done():
-                # tell the caller we are ready
-                ready_future.set_result(tools)
+            initialized = True
             await event.wait()
             logger.info(f"收到 MCP 客户端 {name} 终止信号")
-        except Exception as e:
-            logger.error(f"初始化 MCP 客户端 {name} 失败", exc_info=True)
-            if ready_future and not ready_future.done():
-                ready_future.set_exception(e)
+        except Exception:
+            if not initialized:
+                # 初始化阶段失败，记录错误并向上抛出让 task.exception() 捕获
+                logger.error(f"初始化 MCP 客户端 {name} 失败", exc_info=True)
+                raise
+            # 初始化已成功，此处异常来自 event.wait() 被取消，属于正常终止流程
         finally:
-            # 无论如何都能清理
             await self._terminate_mcp_client(name)
 
     async def _init_mcp_client(self, name: str, config: dict) -> None:

astrbot/core/provider/manager.py

@@ -274,8 +274,8 @@ async def initialize(self):
         if not self.curr_tts_provider_inst and self.tts_provider_insts:
             self.curr_tts_provider_inst = self.tts_provider_insts[0]
 
-        # 初始化 MCP Client 连接
-        asyncio.create_task(self.llm_tools.init_mcp_clients(), name="init_mcp_clients")
+        # 初始化 MCP Client 连接（等待完成以确保工具可用）
+        await self.llm_tools.init_mcp_clients()
 
     def dynamic_import_provider(self, type: str):
         """动态导入提供商适配器模块