[Bug]下载某个插件后出现Cannot read properties of null (reading 'error')

[DDDStar-TheWorld]

What happened / 发生了什么

我在尝试下载并应用某款插件时弹出报错，随后的聊天功能失常，回复错误。在尝试测试模型响应时出现了Cannot read properties of null (reading 'error')

Reproduce / 如何复现？

我的流程为：1.打开astrbot并打开插件市场。
2.下载livingmemory，如出现【加载插「astrbot_plugin_livingmemory」(目录: astrbot_plugin_livingmemory, 版本: 2.2.10) 时出现问题】则为issue流程。
3.测试在astrbot内测试聊天功能或模型响应，聊天功能会回复无内容，复制后显示为【加载插件「astrbot_plugin_livingmemory」(目录: astrbot_plugin_livingmemory, 版本: 2.2.10) 时出现问题】，模型响应则为【Cannot read properties of null (reading 'error')】。
4.删除插件livingmemory后，再次在astrbot测试聊天会出现无规则回复内容，仍然不会直接呈现，在复制后会表现为之前的对话内容，模型响应则仍为【Cannot read properties of null (reading 'error')】。

AstrBot version, deployment method (e.g., Windows Docker Desktop deployment), provider used, and messaging platform used. / AstrBot 版本、部署方式（如 Windows Docker Desktop 部署）、使用的提供商、使用的消息平台适配器

从v4.23.5跟新到v4.23.6，部署方式为 Windows Docker Desktop 部署，使用的提供商有英伟达和openai

OS

Windows

Logs / 报错日志

[2026-04-28 18:12:17.236] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(nvidia/google/gemma-2b) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.244] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(nvidia/google/gemma-2b) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.245] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(nvidia/google/gemma-2b) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.245] [Core] [INFO] [provider.manager:572]: 载入 openai_chat_completion(openai/QC-A02) 服务提供商 ...
[2026-04-28 18:12:17.246] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(openai/QC-A02) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.247] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(openai/QC-A02) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.247] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(openai/QC-A02) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.247] [Core] [INFO] [provider.manager:572]: 载入 openai_chat_completion(openai/QC-A04) 服务提供商 ...
[2026-04-28 18:12:17.248] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(openai/QC-A04) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.249] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(openai/QC-A04) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.249] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(openai/QC-A04) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.250] [Core] [INFO] [provider.manager:572]: 载入 openai_chat_completion(openai/QC-A05-M) 服务提供商 ...
[2026-04-28 18:12:17.250] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(openai/QC-A05-M) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.252] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(openai/QC-A05-M) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.252] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(openai/QC-A05-M) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.252] [Core] [INFO] [provider.manager:567]: Provider openai_1/gpt-5.5 is disabled, skipping
[2026-04-28 18:12:17.253] [Core] [INFO] [provider.manager:567]: Provider openai_1/gpt-5.5-pro is disabled, skipping
[2026-04-28 18:12:17.253] [Core] [INFO] [provider.manager:572]: 载入 openai_chat_completion(openai_1/gpt-5.4-mini) 服务提供商 ...
[2026-04-28 18:12:17.253] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(openai_1/gpt-5.4-mini) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.254] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(openai_1/gpt-5.4-mini) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.254] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(openai_1/gpt-5.4-mini) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.254] [Core] [INFO] [provider.manager:567]: Provider openai_1/gpt-5.4 is disabled, skipping
[2026-04-28 18:12:17.255] [Core] [INFO] [provider.manager:572]: 载入 openai_chat_completion(nvidia_1/deepseek-v4-pro) 服务提供商 ...
[2026-04-28 18:12:17.255] [Core] [ERRO] [v4.23.6] [provider.manager:709]: 实例化 openai_chat_completion(nvidia_1/deepseek-v4-pro) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.256] [Core] [ERRO] [v4.23.6] [provider.manager:283]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 662, in load_provider
inst = cls_type(
^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 281, in initialize
await self.load_provider(provider_config)
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\manager.py", line 712, in load_provider
raise Exception(
Exception: 实例化 openai_chat_completion(nvidia_1/deepseek-v4-pro) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.257] [Core] [ERRO] [v4.23.6] [provider.manager:284]: 实例化 openai_chat_completion(nvidia_1/deepseek-v4-pro) 提供商适配器失败：Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:17.265] [Core] [INFO] [knowledge_base.kb_mgr:39]: 正在初始化知识库模块...
[2026-04-28 18:12:17.292] [Core] [INFO] [knowledge_base.kb_mgr:64]: KnowledgeBase database initialized: C:\Users\xiaon.astrbot\data\knowledge_base\kb.db
[2026-04-28 18:12:17.727] [Core] [INFO] [dashboard.server:391]: 正在启动 WebUI, 监听地址: http://127.0.0.1:6185
[2026-04-28 18:12:19.739] [Core] [INFO] [dashboard.server:430]:
✨✨✨
AstrBot v4.23.6 WebUI 已启动，可访问

➜ 本地: http://localhost:6185
➜ 默认用户名和密码: astrbot
✨✨✨
可在 data/cmd_config.json 中配置 dashboard.host 以便远程访问。
[2026-04-28 18:12:19.740] [Core] [INFO] [core.core_lifecycle:297]: AstrBot 启动完成。
[2026-04-28 18:12:19.765] [Core] [INFO] [utils.temp_dir_cleaner:127]: TempDirCleaner started. interval=600s cleanup_ratio=0.3
[2026-04-28 18:12:19.766] [Core] [INFO] [t2i.network_strategy:84]: Successfully got 3 official T2I endpoints.
[2026-04-28 18:12:20.098] [Core] [INFO] [utils.llm_metadata:63]: Successfully fetched metadata for 2371 LLMs.
[2026-04-28 18:12:49.560] [Core] [INFO] [routes.live_chat:168]: [Live Chat] WebSocket 连接建立: 幽幽晓
[2026-04-28 18:12:49.576] [Core] [INFO] [core.event_bus:61]: [default] [webchat(webchat)] 幽幽晓/幽幽晓: 获取模型列表失败: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:12:49.592] [Core] [WARN] [v4.23.6] [provider.manager:271]: 没有找到 ID 为 openai_1/gpt-5.4-mini 的提供商，这可能是由于您修改了提供商（模型）ID 导致的。
[2026-04-28 18:12:49.592] [Core] [INFO] [core.astr_main_agent:1132]: 未找到任何对话模型（提供商），跳过 LLM 请求处理。
[2026-04-28 18:12:49.598] [Core] [INFO] [routes.live_chat:188]: [Live Chat] WebSocket 连接关闭: 幽幽晓
[2026-04-28 18:15:20.637] [Core] [INFO] [routes.live_chat:168]: [Live Chat] WebSocket 连接建立: 幽幽晓
[2026-04-28 18:15:20.650] [Core] [INFO] [core.event_bus:61]: [default] [webchat(webchat)] 幽幽晓/幽幽晓: 获取模型列表失败: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>
[2026-04-28 18:15:20.664] [Core] [WARN] [v4.23.6] [provider.manager:271]: 没有找到 ID 为 openai_1/gpt-5.4-mini 的提供商，这可能是由于您修改了提供商（模型）ID 导致的。
[2026-04-28 18:15:20.665] [Core] [INFO] [core.astr_main_agent:1132]: 未找到任何对话模型（提供商），跳过 LLM 请求处理。
[2026-04-28 18:15:20.667] [Core] [INFO] [routes.live_chat:188]: [Live Chat] WebSocket 连接关闭: 幽幽晓
[2026-04-28 18:17:19.177] [Core] [INFO] [routes.live_chat:168]: [Live Chat] WebSocket 连接建立: 幽幽晓
[2026-04-28 18:17:19.193] [Core] [INFO] [core.event_bus:61]: [default] [webchat(webchat)] 幽幽晓/幽幽晓: 使用的提供商
[2026-04-28 18:17:19.223] [Core] [WARN] [v4.23.6] [provider.manager:271]: 没有找到 ID 为 openai_1/gpt-5.4-mini 的提供商，这可能是由于您修改了提供商（模型）ID 导致的。
[2026-04-28 18:17:19.223] [Core] [INFO] [core.astr_main_agent:1132]: 未找到任何对话模型（提供商），跳过 LLM 请求处理。
[2026-04-28 18:17:19.226] [Core] [INFO] [routes.live_chat:188]: [Live Chat] WebSocket 连接关闭: 幽幽晓
[2026-04-28 18:17:30.777] [Core] [INFO] [routes.live_chat:168]: [Live Chat] WebSocket 连接建立: 幽幽晓
[2026-04-28 18:17:30.788] [Core] [INFO] [core.event_bus:61]: [default] [webchat(webchat)] 幽幽晓/幽幽晓: 1
[2026-04-28 18:17:30.801] [Core] [WARN] [v4.23.6] [provider.manager:271]: 没有找到 ID 为 openai_1/gpt-5.4-mini 的提供商，这可能是由于您修改了提供商（模型）ID 导致的。
[2026-04-28 18:17:30.801] [Core] [INFO] [core.astr_main_agent:1132]: 未找到任何对话模型（提供商），跳过 LLM 请求处理。
[2026-04-28 18:17:30.805] [Core] [INFO] [routes.live_chat:188]: [Live Chat] WebSocket 连接关闭: 幽幽晓
[2026-04-28 18:17:40.962] [Core] [INFO] [routes.live_chat:168]: [Live Chat] WebSocket 连接建立: 幽幽晓
[2026-04-28 18:17:40.975] [Core] [INFO] [core.event_bus:61]: [default] [webchat(webchat)] 幽幽晓/幽幽晓: 12456785
[2026-04-28 18:17:40.989] [Core] [WARN] [v4.23.6] [provider.manager:271]: 没有找到 ID 为 openai_1/gpt-5.4-mini 的提供商，这可能是由于您修改了提供商（模型）ID 导致的。
[2026-04-28 18:17:40.989] [Core] [INFO] [core.astr_main_agent:1132]: 未找到任何对话模型（提供商），跳过 LLM 请求处理。
[2026-04-28 18:17:40.993] [Core] [INFO] [routes.live_chat:188]: [Live Chat] WebSocket 连接关闭: 幽幽晓
[2026-04-28 18:17:46.648] [Core] [INFO] [routes.plugin:204]: 成功获取远程插件市场数据，包含 1196 个插件
[2026-04-28 18:17:49.222] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai_1/gpt-5.4-mini
[2026-04-28 18:17:49.222] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai_1/gpt-5.4-mini' not found in provider_manager.
[2026-04-28 18:17:49.997] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai_1/gpt-5.4-mini
[2026-04-28 18:17:49.997] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai_1/gpt-5.4-mini' not found in provider_manager.
[2026-04-28 18:19:16.586] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai_1/gpt-5.4-mini
[2026-04-28 18:19:16.587] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai_1/gpt-5.4-mini' not found in provider_manager.
[2026-04-28 18:19:43.989] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai_1/gpt-5.4-mini
[2026-04-28 18:19:43.990] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai_1/gpt-5.4-mini' not found in provider_manager.
[2026-04-28 18:21:23.707] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=nvidia/google/gemma-2b
[2026-04-28 18:21:23.707] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'nvidia/google/gemma-2b' not found in provider_manager.
[2026-04-28 18:21:24.202] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=nvidia/google/gemma-2b
[2026-04-28 18:21:24.202] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'nvidia/google/gemma-2b' not found in provider_manager.
[2026-04-28 18:21:26.298] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=nvidia_1/deepseek-v4-pro
[2026-04-28 18:21:26.299] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'nvidia_1/deepseek-v4-pro' not found in provider_manager.
[2026-04-28 18:21:28.861] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai/QC-A02
[2026-04-28 18:21:28.861] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai/QC-A02' not found in provider_manager.
[2026-04-28 18:21:29.250] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai/QC-A04
[2026-04-28 18:21:29.250] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai/QC-A04' not found in provider_manager.
[2026-04-28 18:21:30.138] [Core] [INFO] [routes.config:743]: API call: /config/provider/check_one id=openai/QC-A05-M
[2026-04-28 18:21:30.138] [Core] [WARN] [v4.23.6] [routes.config:749]: Provider with id 'openai/QC-A05-M' not found in provider_manager.
[2026-04-28 18:21:32.944] [Core] [ERRO] [v4.23.6] [routes.config:1006]: Traceback (most recent call last):
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\dashboard\routes\config.py", line 974, in get_provider_source_models
inst = cls_type(provider_source, {})
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "\?\D:\GitHub\AstrBot\backend\app\astrbot\core\provider\sources\openai_source.py", line 474, in init
self.client = AsyncOpenAI(
^^^^^^^^^^^^
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_client.py", line 617, in init
super().init(
File "D:\GitHub\AstrBot\backend\python\Lib\site-packages\openai_base_client.py", line 1487, in init
raise TypeError(
TypeError: Invalid http_client argument; Expected an instance of httpx.AsyncClient but got <class 'httpx.AsyncClient'>

Are you willing to submit a PR? / 你愿意提交 PR 吗？

• Yes!

Code of Conduct

• I have read and agree to abide by the project's Code of Conduct。

[dosubot]

Hi @DDDStar-TheWorld! I'm Dosu and I’m helping the AstrBot team.

这个问题和 #7770 报告的是同一个根因——插件安装了不同版本的 httpx，导致 OpenAI SDK 内部的 isinstance(http_client, httpx.AsyncClient) 检查失败 [1]。

原因分析： astrbot_plugin_livingmemory（或其依赖）在安装时可能引入了与 AstrBot 主环境不同版本的 httpx。当环境中存在两个 httpx 安装路径时，OpenAI SDK 传入的 httpx.AsyncClient 实例和 SDK 期望的 httpx.AsyncClient 类来自不同路径，isinstance 检查就会失败，即使传入的确实是正确的实例 [2]。

这也解释了为什么删除插件后问题仍然存在——被降级/冲突的 httpx 包不会自动恢复。

临时解决方法：

1. 删除该插件后，进入 Docker 容器手动修复 httpx 版本：

   docker exec -it <容器名> pip install --force-reinstall httpx

2. 然后重启 AstrBot 容器。

目前 Anthropic provider 侧已有类似修复（PR #7685），但 OpenAI provider 尚未应用相同的修复方案 [3]。

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.

[DDDStar-TheWorld]

能否提供详细步骤，我还没尝试过这些类型的修改

[NayukiChiba]

还是pr #7587 和pr #7773 的解决方案，前者是anthropic的模型，后者是openai的模型，等待下版本修复
原因还是httpx的多路径问题