chore: Add PyPI package support for uvx/pip installation (#1764)

* Initial plan

* Add package structure and resource path utilities

- Created langbot/ package with __init__.py and __main__.py entry point
- Added paths utility to find frontend and resource files from package installation
- Updated config loading to use resource paths
- Updated frontend serving to use resource paths
- Added MANIFEST.in for package data inclusion
- Updated pyproject.toml with build system and entry points

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* Add PyPI publishing workflow and update license

- Created GitHub Actions workflow to build frontend and publish to PyPI
- Added license field to pyproject.toml to fix deprecation warning
- Updated .gitignore to exclude build artifacts
- Tested package building successfully

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* Add PyPI installation documentation

- Created PYPI_INSTALLATION.md with detailed installation and usage instructions
- Updated README.md to feature uvx/pip installation as recommended method
- Updated README_EN.md with same changes for English documentation

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* Address code review feedback

- Made package-data configuration more specific to langbot package only
- Improved path detection with caching to avoid repeated file I/O
- Removed sys.path searching which was incorrect for package data
- Removed interactive input() call for non-interactive environment compatibility
- Simplified error messages for version check

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* Fix code review issues

- Use specific exception types instead of bare except
- Fix misleading comments about directory levels
- Remove redundant existence check before makedirs with exist_ok=True
- Use context manager for file opening to ensure proper cleanup

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* Simplify package configuration and document behavioral differences

- Removed redundant package-data configuration, relying on MANIFEST.in
- Added documentation about behavioral differences between package and source installation
- Clarified that include-package-data=true uses MANIFEST.in for data files

Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>

* chore: update pyproject.toml

* chore: try pack templates in langbot/

* chore: update

* chore: update

* chore: update

* chore: update

* chore: update

* chore: adjust dir structure

* chore: fix imports

* fix: read default-pipeline-config.json

* fix: read default-pipeline-config.json

* fix: tests

* ci: publish pypi

* chore: bump version 4.6.0-beta.1 for testing

* chore: add templates/**

* fix: send adapters and requesters icons

* chore: bump version 4.6.0b2 for testing

* chore: add platform field for docker-compose.yaml

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: RockChinQ <45992437+RockChinQ@users.noreply.github.com>
Co-authored-by: Junyan Qin <rockchinq@gmail.com>

src/langbot/pkg/provider/__init__.py

@@ -0,0 +1 @@
+"""OpenAI 接口处理及会话管理相关"""

src/langbot/pkg/provider/modelmgr/entities.py

@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+import typing
+
+import pydantic
+
+from . import requester
+from . import token
+
+
+class LLMModelInfo(pydantic.BaseModel):
+    """模型"""
+
+    name: str
+
+    model_name: typing.Optional[str] = None
+
+    token_mgr: token.TokenManager
+
+    requester: requester.ProviderAPIRequester
+
+    tool_call_supported: typing.Optional[bool] = False
+
+    vision_supported: typing.Optional[bool] = False
+
+    class Config:
+        arbitrary_types_allowed = True

src/langbot/pkg/provider/modelmgr/errors.py

@@ -0,0 +1,5 @@
+class RequesterError(Exception):
+    """Base class for all Requester errors."""
+
+    def __init__(self, message: str):
+        super().__init__('模型请求失败: ' + message)

src/langbot/pkg/provider/modelmgr/modelmgr.py

@@ -0,0 +1,202 @@
+from __future__ import annotations
+
+import sqlalchemy
+import traceback
+
+from . import requester
+from ...core import app
+from ...discover import engine
+from . import token
+from ...entity.persistence import model as persistence_model
+from ...entity.errors import provider as provider_errors
+
+FETCH_MODEL_LIST_URL = 'https://api.qchatgpt.rockchin.top/api/v2/fetch/model_list'
+
+
+class ModelManager:
+    """模型管理器"""
+
+    ap: app.Application
+
+    llm_models: list[requester.RuntimeLLMModel]
+
+    embedding_models: list[requester.RuntimeEmbeddingModel]
+
+    requester_components: list[engine.Component]
+
+    requester_dict: dict[str, type[requester.ProviderAPIRequester]]  # cache
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+        self.llm_models = []
+        self.embedding_models = []
+        self.requester_components = []
+        self.requester_dict = {}
+
+    async def initialize(self):
+        self.requester_components = self.ap.discover.get_components_by_kind('LLMAPIRequester')
+
+        # forge requester class dict
+        requester_dict: dict[str, type[requester.ProviderAPIRequester]] = {}
+        for component in self.requester_components:
+            requester_dict[component.metadata.name] = component.get_python_component_class()
+
+        self.requester_dict = requester_dict
+
+        await self.load_models_from_db()
+
+    async def load_models_from_db(self):
+        """从数据库加载模型"""
+        self.ap.logger.info('Loading models from db...')
+
+        self.llm_models = []
+        self.embedding_models = []
+
+        # llm models
+        result = await self.ap.persistence_mgr.execute_async(sqlalchemy.select(persistence_model.LLMModel))
+        llm_models = result.all()
+        for llm_model in llm_models:
+            try:
+                await self.load_llm_model(llm_model)
+            except provider_errors.RequesterNotFoundError as e:
+                self.ap.logger.warning(f'Requester {e.requester_name} not found, skipping llm model {llm_model.uuid}')
+            except Exception as e:
+                self.ap.logger.error(f'Failed to load model {llm_model.uuid}: {e}\n{traceback.format_exc()}')
+
+        # embedding models
+        result = await self.ap.persistence_mgr.execute_async(sqlalchemy.select(persistence_model.EmbeddingModel))
+        embedding_models = result.all()
+        for embedding_model in embedding_models:
+            try:
+                await self.load_embedding_model(embedding_model)
+            except provider_errors.RequesterNotFoundError as e:
+                self.ap.logger.warning(
+                    f'Requester {e.requester_name} not found, skipping embedding model {embedding_model.uuid}'
+                )
+            except Exception as e:
+                self.ap.logger.error(f'Failed to load model {embedding_model.uuid}: {e}\n{traceback.format_exc()}')
+
+    async def init_runtime_llm_model(
+        self,
+        model_info: persistence_model.LLMModel | sqlalchemy.Row[persistence_model.LLMModel] | dict,
+    ):
+        """初始化运行时 LLM 模型"""
+        if isinstance(model_info, sqlalchemy.Row):
+            model_info = persistence_model.LLMModel(**model_info._mapping)
+        elif isinstance(model_info, dict):
+            model_info = persistence_model.LLMModel(**model_info)
+
+        if model_info.requester not in self.requester_dict:
+            raise provider_errors.RequesterNotFoundError(model_info.requester)
+
+        requester_inst = self.requester_dict[model_info.requester](ap=self.ap, config=model_info.requester_config)
+
+        await requester_inst.initialize()
+
+        runtime_llm_model = requester.RuntimeLLMModel(
+            model_entity=model_info,
+            token_mgr=token.TokenManager(
+                name=model_info.uuid,
+                tokens=model_info.api_keys,
+            ),
+            requester=requester_inst,
+        )
+
+        return runtime_llm_model
+
+    async def init_runtime_embedding_model(
+        self,
+        model_info: persistence_model.EmbeddingModel | sqlalchemy.Row[persistence_model.EmbeddingModel] | dict,
+    ):
+        """初始化运行时 Embedding 模型"""
+        if isinstance(model_info, sqlalchemy.Row):
+            model_info = persistence_model.EmbeddingModel(**model_info._mapping)
+        elif isinstance(model_info, dict):
+            model_info = persistence_model.EmbeddingModel(**model_info)
+
+        if model_info.requester not in self.requester_dict:
+            raise provider_errors.RequesterNotFoundError(model_info.requester)
+
+        requester_inst = self.requester_dict[model_info.requester](ap=self.ap, config=model_info.requester_config)
+
+        await requester_inst.initialize()
+
+        runtime_embedding_model = requester.RuntimeEmbeddingModel(
+            model_entity=model_info,
+            token_mgr=token.TokenManager(
+                name=model_info.uuid,
+                tokens=model_info.api_keys,
+            ),
+            requester=requester_inst,
+        )
+
+        return runtime_embedding_model
+
+    async def load_llm_model(
+        self,
+        model_info: persistence_model.LLMModel | sqlalchemy.Row[persistence_model.LLMModel] | dict,
+    ):
+        """加载 LLM 模型"""
+        runtime_llm_model = await self.init_runtime_llm_model(model_info)
+        self.llm_models.append(runtime_llm_model)
+
+    async def load_embedding_model(
+        self,
+        model_info: persistence_model.EmbeddingModel | sqlalchemy.Row[persistence_model.EmbeddingModel] | dict,
+    ):
+        """加载 Embedding 模型"""
+        runtime_embedding_model = await self.init_runtime_embedding_model(model_info)
+        self.embedding_models.append(runtime_embedding_model)
+
+    async def get_model_by_uuid(self, uuid: str) -> requester.RuntimeLLMModel:
+        """通过uuid获取 LLM 模型"""
+        for model in self.llm_models:
+            if model.model_entity.uuid == uuid:
+                return model
+        raise ValueError(f'LLM model {uuid} not found')
+
+    async def get_embedding_model_by_uuid(self, uuid: str) -> requester.RuntimeEmbeddingModel:
+        """通过uuid获取 Embedding 模型"""
+        for model in self.embedding_models:
+            if model.model_entity.uuid == uuid:
+                return model
+        raise ValueError(f'Embedding model {uuid} not found')
+
+    async def remove_llm_model(self, model_uuid: str):
+        """移除 LLM 模型"""
+        for model in self.llm_models:
+            if model.model_entity.uuid == model_uuid:
+                self.llm_models.remove(model)
+                return
+
+    async def remove_embedding_model(self, model_uuid: str):
+        """移除 Embedding 模型"""
+        for model in self.embedding_models:
+            if model.model_entity.uuid == model_uuid:
+                self.embedding_models.remove(model)
+                return
+
+    def get_available_requesters_info(self, model_type: str) -> list[dict]:
+        """获取所有可用的请求器"""
+        if model_type != '':
+            return [
+                component.to_plain_dict()
+                for component in self.requester_components
+                if model_type in component.spec['support_type']
+            ]
+        else:
+            return [component.to_plain_dict() for component in self.requester_components]
+
+    def get_available_requester_info_by_name(self, name: str) -> dict | None:
+        """通过名称获取请求器信息"""
+        for component in self.requester_components:
+            if component.metadata.name == name:
+                return component.to_plain_dict()
+        return None
+
+    def get_available_requester_manifest_by_name(self, name: str) -> engine.Component | None:
+        """通过名称获取请求器清单"""
+        for component in self.requester_components:
+            if component.metadata.name == name:
+                return component
+        return None

src/langbot/pkg/provider/modelmgr/requester.py

@@ -0,0 +1,142 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from ...core import app
+from ...entity.persistence import model as persistence_model
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+from . import token
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class RuntimeLLMModel:
+    """运行时模型"""
+
+    model_entity: persistence_model.LLMModel
+    """模型数据"""
+
+    token_mgr: token.TokenManager
+    """api key管理器"""
+
+    requester: ProviderAPIRequester
+    """请求器实例"""
+
+    def __init__(
+        self,
+        model_entity: persistence_model.LLMModel,
+        token_mgr: token.TokenManager,
+        requester: ProviderAPIRequester,
+    ):
+        self.model_entity = model_entity
+        self.token_mgr = token_mgr
+        self.requester = requester
+
+
+class RuntimeEmbeddingModel:
+    """运行时 Embedding 模型"""
+
+    model_entity: persistence_model.EmbeddingModel
+    """模型数据"""
+
+    token_mgr: token.TokenManager
+    """api key管理器"""
+
+    requester: ProviderAPIRequester
+    """请求器实例"""
+
+    def __init__(
+        self,
+        model_entity: persistence_model.EmbeddingModel,
+        token_mgr: token.TokenManager,
+        requester: ProviderAPIRequester,
+    ):
+        self.model_entity = model_entity
+        self.token_mgr = token_mgr
+        self.requester = requester
+
+
+class ProviderAPIRequester(metaclass=abc.ABCMeta):
+    """Provider API请求器"""
+
+    name: str = None
+
+    ap: app.Application
+
+    default_config: dict[str, typing.Any] = {}
+
+    requester_cfg: dict[str, typing.Any] = {}
+
+    def __init__(self, ap: app.Application, config: dict[str, typing.Any]):
+        self.ap = ap
+        self.requester_cfg = {**self.default_config}
+        self.requester_cfg.update(config)
+
+    async def initialize(self):
+        pass
+
+    @abc.abstractmethod
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        """调用API
+
+        Args:
+            model (RuntimeLLMModel): 使用的模型信息
+            messages (typing.List[llm_entities.Message]): 消息对象列表
+            funcs (typing.List[tools_entities.LLMFunction], optional): 使用的工具函数列表. Defaults to None.
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+            remove_think (bool, optional): 是否移思考中的消息. Defaults to False.
+
+        Returns:
+            llm_entities.Message: 返回消息对象
+        """
+        pass
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        """调用API
+
+        Args:
+            model (RuntimeLLMModel): 使用的模型信息
+            messages (typing.List[provider_message.Message]): 消息对象列表
+            funcs (typing.List[resource_tool.LLMTool], optional): 使用的工具函数列表. Defaults to None.
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+            remove_think (bool, optional): 是否移除思考中的消息. Defaults to False.
+
+        Returns:
+            typing.AsyncGenerator[provider_message.MessageChunk]: 返回消息对象
+        """
+        pass
+
+    async def invoke_embedding(
+        self,
+        model: RuntimeEmbeddingModel,
+        input_text: typing.List[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> typing.List[typing.List[float]]:
+        """调用 Embedding API
+
+        Args:
+            model (RuntimeEmbeddingModel): 使用的模型信息
+            input_text (typing.List[str]): 输入文本
+            extra_args (dict[str, typing.Any], optional): 额外的参数. Defaults to {}.
+
+        Returns:
+            typing.List[typing.List[float]]: 返回的 embedding 向量
+        """
+        pass

src/langbot/pkg/provider/modelmgr/requester.yaml

@@ -0,0 +1,14 @@
+apiVersion: v1
+kind: ComponentTemplate
+metadata:
+  name: LLMAPIRequester
+  label:
+    en_US: LLM API Requester
+    zh_Hans: LLM API 请求器
+spec:
+  type:
+    - python
+execution:
+  python:
+    path: ./requester.py
+    attr: LLMAPIRequester

src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class AI302ChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """302.AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.302.ai/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/302aichatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: 302-ai-chat-completions
+  label:
+    en_US: 302.AI
+    zh_Hans: 302.AI
+  icon: 302ai.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.302.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./302aichatcmpl.py
+    attr: AI302ChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/anthropic.svg

@@ -0,0 +1,4 @@
+<svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg">
+<rect width="24" height="24" rx="6" fill="#CA9F7B"/>
+<path d="M15.3843 6.43481H12.9687L17.3739 17.5652H19.7896L15.3843 6.43481ZM8.40522 6.43481L4 17.5652H6.4633L7.36417 15.2279H11.9729L12.8737 17.5652H15.337L10.9318 6.43481H8.40522ZM8.16104 13.1607L9.66852 9.24907L11.176 13.1607H8.16104Z" fill="#191918"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.py

@@ -0,0 +1,370 @@
+from __future__ import annotations
+
+import typing
+import json
+import platform
+import socket
+import anthropic
+import httpx
+
+from .. import errors, requester
+
+from ....utils import image
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class AnthropicMessages(requester.ProviderAPIRequester):
+    """Anthropic Messages API 请求器"""
+
+    client: anthropic.AsyncAnthropic
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.anthropic.com',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        # 兼容 Windows 缺失 TCP_KEEPINTVL 和 TCP_KEEPCNT 的问题
+        if platform.system() == 'Windows':
+            if not hasattr(socket, 'TCP_KEEPINTVL'):
+                socket.TCP_KEEPINTVL = 0
+            if not hasattr(socket, 'TCP_KEEPCNT'):
+                socket.TCP_KEEPCNT = 0
+        httpx_client = anthropic._base_client.AsyncHttpxClientWrapper(
+            base_url=self.requester_cfg['base_url'],
+            # cast to a valid type because mypy doesn't understand our type narrowing
+            timeout=typing.cast(httpx.Timeout, self.requester_cfg['timeout']),
+            limits=anthropic._constants.DEFAULT_CONNECTION_LIMITS,
+            follow_redirects=True,
+            trust_env=True,
+        )
+
+        self.client = anthropic.AsyncAnthropic(
+            api_key='',
+            http_client=httpx_client,
+            base_url=self.requester_cfg['base_url'],
+        )
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = extra_args.copy()
+        args['model'] = model.model_entity.name
+
+        # 处理消息
+
+        # system
+        system_role_message = None
+
+        for i, m in enumerate(messages):
+            if m.role == 'system':
+                system_role_message = m
+
+                break
+
+        if system_role_message:
+            messages.pop(i)
+
+        if isinstance(system_role_message, provider_message.Message) and isinstance(system_role_message.content, str):
+            args['system'] = system_role_message.content
+
+        req_messages = []
+
+        for m in messages:
+            if m.role == 'tool':
+                tool_call_id = m.tool_call_id
+
+                req_messages.append(
+                    {
+                        'role': 'user',
+                        'content': [
+                            {
+                                'type': 'tool_result',
+                                'tool_use_id': tool_call_id,
+                                'is_error': False,
+                                'content': [{'type': 'text', 'text': m.content}],
+                            }
+                        ],
+                    }
+                )
+
+                continue
+
+            msg_dict = m.dict(exclude_none=True)
+
+            if isinstance(m.content, str) and m.content.strip() != '':
+                msg_dict['content'] = [{'type': 'text', 'text': m.content}]
+            elif isinstance(m.content, list):
+                for i, ce in enumerate(m.content):
+                    if ce.type == 'image_base64':
+                        image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+
+                        alter_image_ele = {
+                            'type': 'image',
+                            'source': {
+                                'type': 'base64',
+                                'media_type': f'image/{image_format}',
+                                'data': image_b64,
+                            },
+                        }
+                        msg_dict['content'][i] = alter_image_ele
+
+            if m.tool_calls:
+                for tool_call in m.tool_calls:
+                    msg_dict['content'].append(
+                        {
+                            'type': 'tool_use',
+                            'id': tool_call.id,
+                            'name': tool_call.function.name,
+                            'input': json.loads(tool_call.function.arguments),
+                        }
+                    )
+
+                del msg_dict['tool_calls']
+
+            req_messages.append(msg_dict)
+
+        args['messages'] = req_messages
+
+        if 'thinking' in args:
+            args['thinking'] = {'type': 'enabled', 'budget_tokens': 10000}
+
+        if funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_anthropic(funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        try:
+            resp = await self.client.messages.create(**args)
+
+            args = {
+                'content': '',
+                'role': resp.role,
+            }
+            assert type(resp) is anthropic.types.message.Message
+
+            for block in resp.content:
+                if not remove_think and block.type == 'thinking':
+                    args['content'] = '<think>\n' + block.thinking + '\n</think>\n' + args['content']
+                elif block.type == 'text':
+                    args['content'] += block.text
+                elif block.type == 'tool_use':
+                    assert type(block) is anthropic.types.tool_use_block.ToolUseBlock
+                    tool_call = provider_message.ToolCall(
+                        id=block.id,
+                        type='function',
+                        function=provider_message.FunctionCall(name=block.name, arguments=json.dumps(block.input)),
+                    )
+                    if 'tool_calls' not in args:
+                        args['tool_calls'] = []
+                    args['tool_calls'].append(tool_call)
+
+            return provider_message.Message(**args)
+        except anthropic.AuthenticationError as e:
+            raise errors.RequesterError(f'api-key 无效: {e.message}')
+        except anthropic.BadRequestError as e:
+            raise errors.RequesterError(str(e.message))
+        except anthropic.NotFoundError as e:
+            if 'model: ' in str(e):
+                raise errors.RequesterError(f'模型无效: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求地址无效: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = extra_args.copy()
+        args['model'] = model.model_entity.name
+        args['stream'] = True
+
+        # 处理消息
+
+        # system
+        system_role_message = None
+
+        for i, m in enumerate(messages):
+            if m.role == 'system':
+                system_role_message = m
+
+                break
+
+        if system_role_message:
+            messages.pop(i)
+
+        if isinstance(system_role_message, provider_message.Message) and isinstance(system_role_message.content, str):
+            args['system'] = system_role_message.content
+
+        req_messages = []
+
+        for m in messages:
+            if m.role == 'tool':
+                tool_call_id = m.tool_call_id
+
+                req_messages.append(
+                    {
+                        'role': 'user',
+                        'content': [
+                            {
+                                'type': 'tool_result',
+                                'tool_use_id': tool_call_id,
+                                'is_error': False,  # 暂时直接写false
+                                'content': [
+                                    {'type': 'text', 'text': m.content}
+                                ],  # 这里要是list包裹，应该是多个返回的情况？type类型好像也可以填其他的，暂时只写text
+                            }
+                        ],
+                    }
+                )
+
+                continue
+
+            msg_dict = m.dict(exclude_none=True)
+
+            if isinstance(m.content, str) and m.content.strip() != '':
+                msg_dict['content'] = [{'type': 'text', 'text': m.content}]
+            elif isinstance(m.content, list):
+                for i, ce in enumerate(m.content):
+                    if ce.type == 'image_base64':
+                        image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+
+                        alter_image_ele = {
+                            'type': 'image',
+                            'source': {
+                                'type': 'base64',
+                                'media_type': f'image/{image_format}',
+                                'data': image_b64,
+                            },
+                        }
+                        msg_dict['content'][i] = alter_image_ele
+            if isinstance(msg_dict['content'], str) and msg_dict['content'] == '':
+                msg_dict['content'] = []  # 这里不知道为什么会莫名有个空导致content为字符
+            if m.tool_calls:
+                for tool_call in m.tool_calls:
+                    msg_dict['content'].append(
+                        {
+                            'type': 'tool_use',
+                            'id': tool_call.id,
+                            'name': tool_call.function.name,
+                            'input': json.loads(tool_call.function.arguments),
+                        }
+                    )
+
+                del msg_dict['tool_calls']
+
+            req_messages.append(msg_dict)
+        if 'thinking' in args:
+            args['thinking'] = {'type': 'enabled', 'budget_tokens': 10000}
+
+        args['messages'] = req_messages
+
+        if funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_anthropic(funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        try:
+            role = 'assistant'  # 默认角色
+            # chunk_idx = 0
+            think_started = False
+            think_ended = False
+            finish_reason = False
+            content = ''
+            tool_name = ''
+            tool_id = ''
+            async for chunk in await self.client.messages.create(**args):
+                tool_call = {'id': None, 'function': {'name': None, 'arguments': None}, 'type': 'function'}
+                if isinstance(
+                    chunk, anthropic.types.raw_content_block_start_event.RawContentBlockStartEvent
+                ):  # 记录开始
+                    if chunk.content_block.type == 'tool_use':
+                        if chunk.content_block.name is not None:
+                            tool_name = chunk.content_block.name
+                        if chunk.content_block.id is not None:
+                            tool_id = chunk.content_block.id
+
+                        tool_call['function']['name'] = tool_name
+                        tool_call['function']['arguments'] = ''
+                        tool_call['id'] = tool_id
+
+                    if not remove_think:
+                        if chunk.content_block.type == 'thinking' and not remove_think:
+                            think_started = True
+                        elif chunk.content_block.type == 'text' and chunk.index != 0 and not remove_think:
+                            think_ended = True
+                        continue
+                elif isinstance(chunk, anthropic.types.raw_content_block_delta_event.RawContentBlockDeltaEvent):
+                    if chunk.delta.type == 'thinking_delta':
+                        if think_started:
+                            think_started = False
+                            content = '<think>\n' + chunk.delta.thinking
+                        elif remove_think:
+                            continue
+                        else:
+                            content = chunk.delta.thinking
+                    elif chunk.delta.type == 'text_delta':
+                        if think_ended:
+                            think_ended = False
+                            content = '\n</think>\n' + chunk.delta.text
+                        else:
+                            content = chunk.delta.text
+                    elif chunk.delta.type == 'input_json_delta':
+                        tool_call['function']['arguments'] = chunk.delta.partial_json
+                        tool_call['function']['name'] = tool_name
+                        tool_call['id'] = tool_id
+                elif isinstance(chunk, anthropic.types.raw_content_block_stop_event.RawContentBlockStopEvent):
+                    continue  # 记录raw_content_block结束的
+
+                elif isinstance(chunk, anthropic.types.raw_message_delta_event.RawMessageDeltaEvent):
+                    if chunk.delta.stop_reason == 'end_turn':
+                        finish_reason = True
+                elif isinstance(chunk, anthropic.types.raw_message_stop_event.RawMessageStopEvent):
+                    continue  # 这个好像是完全结束
+                else:
+                    # print(chunk)
+                    self.ap.logger.debug(f'anthropic chunk: {chunk}')
+                    continue
+
+                args = {
+                    'content': content,
+                    'role': role,
+                    'is_final': finish_reason,
+                    'tool_calls': None if tool_call['id'] is None else [tool_call],
+                }
+                # if chunk_idx == 0:
+                #     chunk_idx += 1
+                #     continue
+
+                # assert type(chunk) is anthropic.types.message.Chunk
+
+                yield provider_message.MessageChunk(**args)
+
+            # return llm_entities.Message(**args)
+        except anthropic.AuthenticationError as e:
+            raise errors.RequesterError(f'api-key 无效: {e.message}')
+        except anthropic.BadRequestError as e:
+            raise errors.RequesterError(str(e.message))
+        except anthropic.NotFoundError as e:
+            if 'model: ' in str(e):
+                raise errors.RequesterError(f'模型无效: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求地址无效: {e.message}')

src/langbot/pkg/provider/modelmgr/requesters/anthropicmsgs.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: anthropic-messages
+  label:
+    en_US: Anthropic
+    zh_Hans: Anthropic
+  icon: anthropic.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.anthropic.com
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./anthropicmsgs.py
+    attr: AnthropicMessages

src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.py

@@ -0,0 +1,242 @@
+from __future__ import annotations
+
+import typing
+import dashscope
+import openai
+
+from . import modelscopechatcmpl
+from .. import requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class BailianChatCompletions(modelscopechatcmpl.ModelScopeChatCompletions):
+    """阿里云百炼大模型平台 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://dashscope.aliyuncs.com/compatible-mode/v1',
+        'timeout': 120,
+    }
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        is_use_dashscope_call = False  # 是否使用阿里原生库调用
+        is_enable_multi_model = True  # 是否支持多轮对话
+        use_time_num = 0  # 模型已调用次数，防止存在多文件时重复调用
+        use_time_ids = []  # 已调用的ID列表
+        message_id = 0  # 记录消息序号
+
+        for msg in messages:
+            # print(msg)
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+                    elif me['type'] == 'file_url' and '.' in me.get('file_name', ''):
+                        # 1. 视频文件推理
+                        # https://bailian.console.aliyun.com/?tab=doc#/doc/?type=model&url=2845871
+                        file_type = me.get('file_name').lower().split('.')[-1]
+                        if file_type in ['mp4', 'avi', 'mkv', 'mov', 'flv', 'wmv']:
+                            me['type'] = 'video_url'
+                            me['video_url'] = {'url': me['file_url']}
+                            del me['file_url']
+                            del me['file_name']
+                            use_time_num += 1
+                            use_time_ids.append(message_id)
+                            is_enable_multi_model = False
+                        # 2. 语音文件识别, 无法通过openai的audio字段传递，暂时不支持
+                        # https://bailian.console.aliyun.com/?tab=doc#/doc/?type=model&url=2979031
+                        elif file_type in [
+                            'aac',
+                            'amr',
+                            'aiff',
+                            'flac',
+                            'm4a',
+                            'mp3',
+                            'mpeg',
+                            'ogg',
+                            'opus',
+                            'wav',
+                            'webm',
+                            'wma',
+                        ]:
+                            me['audio'] = me['file_url']
+                            me['type'] = 'audio'
+                            del me['file_url']
+                            del me['type']
+                            del me['file_name']
+                            is_use_dashscope_call = True
+                            use_time_num += 1
+                            use_time_ids.append(message_id)
+                            is_enable_multi_model = False
+            message_id += 1
+
+        # 使用列表推导式，保留不在 use_time_ids[:-1] 中的元素，仅保留最后一个多媒体消息
+        if not is_enable_multi_model and use_time_num > 1:
+            messages = [msg for idx, msg in enumerate(messages) if idx not in use_time_ids[:-1]]
+
+        if not is_enable_multi_model:
+            messages = [msg for msg in messages if 'resp_message_id' not in msg]
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+
+        if is_use_dashscope_call:
+            response = dashscope.MultiModalConversation.call(
+                # 若没有配置环境变量，请用百炼API Key将下行替换为：api_key = "sk-xxx"
+                api_key=use_model.token_mgr.get_token(),
+                model=use_model.model_entity.name,
+                messages=messages,
+                result_format='message',
+                asr_options={
+                    # "language": "zh", # 可选，若已知音频的语种，可通过该参数指定待识别语种，以提升识别准确率
+                    'enable_lid': True,
+                    'enable_itn': False,
+                },
+                stream=True,
+            )
+            content_length_list = []
+            previous_length = 0  # 记录上一次的内容长度
+            for res in response:
+                chunk = res['output']
+                # 解析 chunk 数据
+                if hasattr(chunk, 'choices') and chunk.choices:
+                    choice = chunk.choices[0]
+                    delta_content = choice['message'].content[0]['text']
+                    finish_reason = choice['finish_reason']
+                    content_length_list.append(len(delta_content))
+                else:
+                    delta_content = ''
+                    finish_reason = None
+
+                # 跳过空的第一个 chunk（只有 role 没有内容）
+                if chunk_idx == 0 and not delta_content:
+                    chunk_idx += 1
+                    continue
+
+                # 检查 content_length_list 是否有足够的数据
+                if len(content_length_list) >= 2:
+                    now_content = delta_content[previous_length : content_length_list[-1]]
+                    previous_length = content_length_list[-1]  # 更新上一次的长度
+                else:
+                    now_content = delta_content  # 第一次循环时直接使用 delta_content
+                    previous_length = len(delta_content)  # 更新上一次的长度
+
+                # 构建 MessageChunk - 只包含增量内容
+                chunk_data = {
+                    'role': role,
+                    'content': now_content if now_content else None,
+                    'is_final': bool(finish_reason) and finish_reason != 'null',
+                }
+
+                # 移除 None 值
+                chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+                yield provider_message.MessageChunk(**chunk_data)
+                chunk_idx += 1
+        else:
+            async for chunk in self._req_stream(args, extra_body=extra_args):
+                # 解析 chunk 数据
+                if hasattr(chunk, 'choices') and chunk.choices:
+                    choice = chunk.choices[0]
+                    delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                    finish_reason = getattr(choice, 'finish_reason', None)
+                else:
+                    delta = {}
+                    finish_reason = None
+
+                # 从第一个 chunk 获取 role，后续使用这个 role
+                if 'role' in delta and delta['role']:
+                    role = delta['role']
+
+                # 获取增量内容
+                delta_content = delta.get('content', '')
+                reasoning_content = delta.get('reasoning_content', '')
+
+                # 处理 reasoning_content
+                if reasoning_content:
+                    # accumulated_reasoning += reasoning_content
+                    # 如果设置了 remove_think，跳过 reasoning_content
+                    if remove_think:
+                        chunk_idx += 1
+                        continue
+
+                    # 第一次出现 reasoning_content，添加 <think> 开始标签
+                    if not thinking_started:
+                        thinking_started = True
+                        delta_content = '<think>\n' + reasoning_content
+                    else:
+                        # 继续输出 reasoning_content
+                        delta_content = reasoning_content
+                elif thinking_started and not thinking_ended and delta_content:
+                    # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                    thinking_ended = True
+                    delta_content = '\n</think>\n' + delta_content
+
+                # 处理工具调用增量
+                if delta.get('tool_calls'):
+                    for tool_call in delta['tool_calls']:
+                        if tool_call['id'] != '':
+                            tool_id = tool_call['id']
+                        if tool_call['function']['name'] is not None:
+                            tool_name = tool_call['function']['name']
+
+                        if tool_call['type'] is None:
+                            tool_call['type'] = 'function'
+                        tool_call['id'] = tool_id
+                        tool_call['function']['name'] = tool_name
+                        tool_call['function']['arguments'] = (
+                            '' if tool_call['function']['arguments'] is None else tool_call['function']['arguments']
+                        )
+
+                # 跳过空的第一个 chunk（只有 role 没有内容）
+                if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                    chunk_idx += 1
+                    continue
+
+                # 构建 MessageChunk - 只包含增量内容
+                chunk_data = {
+                    'role': role,
+                    'content': delta_content if delta_content else None,
+                    'tool_calls': delta.get('tool_calls'),
+                    'is_final': bool(finish_reason),
+                }
+
+                # 移除 None 值
+                chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+                yield provider_message.MessageChunk(**chunk_data)
+                chunk_idx += 1
+                # return

src/langbot/pkg/provider/modelmgr/requesters/bailianchatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: bailian-chat-completions
+  label:
+    en_US: Aliyun Bailian
+    zh_Hans: 阿里云百炼
+  icon: bailian.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://dashscope.aliyuncs.com/compatible-mode/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./bailianchatcmpl.py
+    attr: BailianChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.py

@@ -0,0 +1,406 @@
+from __future__ import annotations
+
+import asyncio
+import typing
+
+import openai
+import openai.types.chat.chat_completion as chat_completion
+import httpx
+
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class OpenAIChatCompletions(requester.ProviderAPIRequester):
+    """OpenAI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.openai.com/v1',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        self.client = openai.AsyncClient(
+            api_key='',
+            base_url=self.requester_cfg['base_url'].replace(' ', ''),
+            timeout=self.requester_cfg['timeout'],
+            http_client=httpx.AsyncClient(trust_env=True, timeout=self.requester_cfg['timeout']),
+        )
+
+    async def _req(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        return await self.client.chat.completions.create(**args, extra_body=extra_body)
+
+    async def _req_stream(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ):
+        async for chunk in await self.client.chat.completions.create(**args, extra_body=extra_body):
+            yield chunk
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        # 处理思维链
+        content = chatcmpl_message.get('content', '')
+        reasoning_content = chatcmpl_message.get('reasoning_content', None)
+
+        processed_content, _ = await self._process_thinking_content(
+            content=content, reasoning_content=reasoning_content, remove_think=remove_think
+        )
+
+        chatcmpl_message['content'] = processed_content
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            (处理后的内容, 提取的思维链内容)
+        """
+        thinking_content = ''
+
+        # 1. 从 reasoning_content 提取思维链
+        if reasoning_content:
+            thinking_content = reasoning_content
+
+        # 2. 从 content 中提取 <think> 标签内容
+        if content and '<think>' in content and '</think>' in content:
+            import re
+
+            think_pattern = r'<think>(.*?)</think>'
+            think_matches = re.findall(think_pattern, content, re.DOTALL)
+            if think_matches:
+                # 如果已有 reasoning_content，则追加
+                if thinking_content:
+                    thinking_content += '\n' + '\n'.join(think_matches)
+                else:
+                    thinking_content = '\n'.join(think_matches)
+                # 移除 content 中的 <think> 标签
+                content = re.sub(think_pattern, '', content, flags=re.DOTALL).strip()
+
+        # 3. 根据 remove_think 参数决定是否保留思维链
+        if remove_think:
+            return content, ''
+        else:
+            # 如果有思维链内容，将其以 <think> 格式添加到 content 开头
+            if thinking_content:
+                content = f'<think>\n{thinking_content}\n</think>\n{content}'.strip()
+            return content, thinking_content
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        tool_id = ''
+        tool_name = ''
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+                    else:
+                        tool_call['id'] = tool_id
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+
+        # 发送请求
+
+        resp = await self._req(args, extra_body=extra_args)
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            msg = await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+            return msg
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
+
+    async def invoke_embedding(
+        self,
+        model: requester.RuntimeEmbeddingModel,
+        input_text: list[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> list[list[float]]:
+        """调用 Embedding API"""
+        self.client.api_key = model.token_mgr.get_token()
+
+        args = {
+            'model': model.model_entity.name,
+            'input': input_text,
+        }
+
+        if model.model_entity.extra_args:
+            args.update(model.model_entity.extra_args)
+
+        args.update(extra_args)
+
+        try:
+            resp = await self.client.embeddings.create(**args)
+
+            return [d.embedding for d in resp.data]
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            raise errors.RequesterError(f'请求参数错误: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            async for item in self._closure_stream(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            ):
+                yield item
+
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')

src/langbot/pkg/provider/modelmgr/requesters/chatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: openai-chat-completions
+  label:
+    en_US: OpenAI
+    zh_Hans: OpenAI
+  icon: openai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.openai.com/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./chatcmpl.py
+    attr: OpenAIChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class CompShareChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """CompShare ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.modelverse.cn/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/compsharechatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: compshare-chat-completions
+  label:
+    en_US: CompShare
+    zh_Hans: 优云智算
+  icon: compshare.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.modelverse.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./compsharechatcmpl.py
+    attr: CompShareChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/deepseek.svg

@@ -0,0 +1,3 @@
+<svg width="60" height="50" viewBox="0 0 60 50" version="1.1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" xml:space="preserve" xmlns:serif="http://www.serif.com/" style="fill-rule:evenodd;clip-rule:evenodd;stroke-linejoin:round;stroke-miterlimit:2;">
+    <path id="path" d="M55.613,3.471C55.018,3.179 54.761,3.735 54.413,4.018C54.294,4.109 54.193,4.227 54.092,4.337C53.222,5.266 52.206,5.876 50.878,5.803C48.936,5.694 47.278,6.305 45.813,7.79C45.501,5.959 44.466,4.865 42.891,4.164C42.067,3.799 41.233,3.435 40.656,2.642C40.253,2.077 40.143,1.448 39.942,0.829C39.814,0.455 39.685,0.073 39.255,0.009C38.788,-0.064 38.605,0.328 38.421,0.656C37.689,1.995 37.405,3.471 37.432,4.965C37.496,8.327 38.916,11.006 41.737,12.91C42.058,13.129 42.14,13.347 42.039,13.666C41.847,14.323 41.618,14.96 41.416,15.616C41.288,16.035 41.096,16.127 40.647,15.944C39.099,15.297 37.762,14.341 36.58,13.184C34.575,11.243 32.761,9.102 30.499,7.425C29.968,7.033 29.436,6.669 28.887,6.323C26.579,4.082 29.189,2.241 29.794,2.022C30.425,1.795 30.013,1.011 27.971,1.02C25.928,1.029 24.06,1.713 21.679,2.624C21.331,2.761 20.964,2.861 20.589,2.943C18.427,2.533 16.183,2.441 13.838,2.706C9.424,3.198 5.898,5.284 3.306,8.847C0.191,13.129 -0.541,17.994 0.356,23.069C1.3,28.417 4.029,32.845 8.224,36.308C12.575,39.897 17.584,41.656 23.3,41.319C26.771,41.118 30.636,40.654 34.996,36.964C36.095,37.51 37.249,37.729 39.163,37.893C40.638,38.03 42.058,37.82 43.157,37.592C44.878,37.228 44.759,35.633 44.137,35.342C39.09,32.991 40.198,33.948 39.191,33.173C41.755,30.139 45.62,26.987 47.132,16.773C47.251,15.962 47.15,15.452 47.132,14.796C47.122,14.395 47.214,14.241 47.672,14.195C48.936,14.049 50.163,13.703 51.29,13.083C54.56,11.298 55.878,8.364 56.19,4.847C56.236,4.309 56.181,3.753 55.613,3.471ZM27.119,35.123C22.228,31.278 19.856,30.012 18.876,30.066C17.96,30.121 18.125,31.169 18.326,31.852C18.537,32.526 18.812,32.991 19.196,33.583C19.462,33.975 19.645,34.558 18.931,34.996C17.356,35.971 14.617,34.667 14.489,34.604C11.301,32.727 8.636,30.249 6.758,26.859C4.945,23.597 3.892,20.099 3.718,16.363C3.672,15.461 3.938,15.142 4.835,14.979C6.017,14.76 7.235,14.714 8.416,14.887C13.408,15.616 17.658,17.848 21.221,21.384C23.254,23.397 24.793,25.802 26.377,28.153C28.062,30.649 29.876,33.028 32.184,34.977C32.999,35.661 33.649,36.18 34.272,36.563C32.395,36.772 29.262,36.818 27.119,35.123ZM29.464,20.044C29.464,19.643 29.784,19.325 30.187,19.325C30.279,19.325 30.361,19.343 30.435,19.37C30.535,19.407 30.627,19.461 30.7,19.543C30.828,19.671 30.902,19.853 30.902,20.044C30.902,20.445 30.581,20.764 30.178,20.764C29.775,20.764 29.464,20.445 29.464,20.044ZM36.745,23.78C36.278,23.971 35.811,24.135 35.362,24.153C34.666,24.19 33.906,23.907 33.494,23.561C32.853,23.023 32.395,22.723 32.202,21.784C32.12,21.384 32.166,20.764 32.239,20.409C32.404,19.643 32.221,19.151 31.68,18.705C31.241,18.34 30.682,18.24 30.068,18.24C29.839,18.24 29.629,18.14 29.473,18.058C29.216,17.93 29.006,17.612 29.207,17.22C29.271,17.092 29.583,16.783 29.656,16.728C30.49,16.254 31.451,16.409 32.34,16.764C33.164,17.101 33.787,17.721 34.684,18.596C35.6,19.652 35.765,19.944 36.287,20.737C36.7,21.356 37.075,21.994 37.331,22.723C37.487,23.179 37.286,23.552 36.745,23.78Z" style="fill:rgb(77,107,254);fill-rule:nonzero;"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.py

@@ -0,0 +1,60 @@
+from __future__ import annotations
+
+import typing
+
+from . import chatcmpl
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class DeepseekChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Deepseek ChatCompletion API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.deepseek.com',
+        'timeout': 120,
+    }
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages
+
+        # deepseek 不支持多模态，把content都转换成纯文字
+        for m in messages:
+            if 'content' in m and isinstance(m['content'], list):
+                m['content'] = ' '.join([c['text'] for c in m['content']])
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(args, extra_body=extra_args)
+
+        # print(resp)
+
+        if resp is None:
+            raise errors.RequesterError('接口返回为空，请确定模型提供商服务是否正常')
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message

src/langbot/pkg/provider/modelmgr/requesters/deepseekchatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: deepseek-chat-completions
+  label:
+    en_US: DeepSeek
+    zh_Hans: DeepSeek
+  icon: deepseek.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.deepseek.com
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./deepseekchatcmpl.py
+    attr: DeepseekChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/gemini.svg

@@ -0,0 +1 @@
+<svg height="1em" style="flex:none;line-height:1" viewBox="0 0 24 24" width="1em" xmlns="http://www.w3.org/2000/svg"><title>Gemini</title><defs><linearGradient id="lobe-icons-gemini-fill" x1="0%" x2="68.73%" y1="100%" y2="30.395%"><stop offset="0%" stop-color="#1C7DFF"></stop><stop offset="52.021%" stop-color="#1C69FF"></stop><stop offset="100%" stop-color="#F0DCD6"></stop></linearGradient></defs><path d="M12 24A14.304 14.304 0 000 12 14.304 14.304 0 0012 0a14.305 14.305 0 0012 12 14.305 14.305 0 00-12 12" fill="url(#lobe-icons-gemini-fill)" fill-rule="nonzero"></path></svg>

src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.py

@@ -0,0 +1,142 @@
+from __future__ import annotations
+
+import typing
+
+from . import chatcmpl
+
+import uuid
+
+from .. import requester
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class GeminiChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Google Gemini API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://generativelanguage.googleapis.com/v1beta/openai',
+        'timeout': 120,
+    }
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        tool_id = ''
+        tool_name = ''
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] == '' and tool_id == '':
+                        tool_id = str(uuid.uuid4())
+                    if tool_call['function']['name']:
+                        tool_name = tool_call['function']['name']
+                    tool_call['id'] = tool_id
+                    tool_call['function']['name'] = tool_name
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1

src/langbot/pkg/provider/modelmgr/requesters/geminichatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: gemini-chat-completions
+  label:
+    en_US: Google Gemini
+    zh_Hans: Google Gemini
+  icon: gemini.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://generativelanguage.googleapis.com/v1beta/openai
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./geminichatcmpl.py
+    attr: GeminiChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/giteeai.svg

@@ -0,0 +1,3 @@
+<svg width="40" height="40" viewBox="0 0 40 40" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path fill-rule="evenodd" clip-rule="evenodd" d="M25.132 24.3947C25.497 25.7527 25.8984 27.1413 26.3334 28.5834C26.7302 29.8992 25.5459 30.4167 25.0752 29.1758C24.571 27.8466 24.0885 26.523 23.6347 25.1729C21.065 26.4654 18.5025 27.5424 15.5961 28.7541C16.7581 33.0256 17.8309 36.5984 19.4952 39.9935C19.4953 39.9936 19.4953 39.9937 19.4954 39.9938C19.6631 39.9979 19.8313 40 20 40C31.0457 40 40 31.0457 40 20C40 16.0335 38.8453 12.3366 36.8537 9.22729C31.6585 9.69534 27.0513 10.4562 22.8185 11.406C22.8882 12.252 22.9677 13.0739 23.0555 13.855C23.3824 16.7604 23.9112 19.5281 24.6137 22.3836C27.0581 21.2848 29.084 20.3225 30.6816 19.522C32.2154 18.7535 33.6943 18.7062 31.2018 20.6594C29.0388 22.1602 27.0644 23.3566 25.132 24.3947ZM36.1559 8.20846C33.0001 3.89184 28.1561 0.887462 22.5955 0.166882C22.4257 2.86234 22.4785 6.26344 22.681 9.50447C26.7473 8.88859 31.1721 8.46032 36.1559 8.20846ZM19.9369 9.73661e-05C19.7594 2.92694 19.8384 6.65663 20.19 9.91293C17.3748 10.4109 14.7225 11.0064 12.1592 11.7038C12.0486 10.4257 11.9927 9.25764 11.9927 8.24178C11.9927 7.5054 11.3957 6.90844 10.6593 6.90844C9.92296 6.90844 9.32601 7.5054 9.32601 8.24178C9.32601 9.47868 9.42873 10.898 9.61402 12.438C8.33567 12.8278 7.07397 13.2443 5.81918 13.688C5.12493 13.9336 4.76118 14.6954 5.0067 15.3896C5.25223 16.0839 6.01406 16.4476 6.7083 16.2021C7.7931 15.8185 8.88482 15.4388 9.98927 15.0659C10.5222 18.3344 11.3344 21.9428 12.2703 25.4156C12.4336 26.0218 12.6062 26.6262 12.7863 27.2263C9.34168 28.4135 5.82612 29.3782 2.61128 29.8879C0.949407 26.9716 0 23.5967 0 20C0 8.97534 8.92023 0.0341108 19.9369 9.73661e-05ZM4.19152 32.2527C7.45069 36.4516 12.3458 39.3173 17.9204 39.8932C16.5916 37.455 14.9338 33.717 13.5405 29.5901C10.4404 30.7762 7.25883 31.6027 4.19152 32.2527ZM22.9735 23.1135C22.1479 20.41 21.4462 17.5441 20.9225 14.277C20.746 13.5841 20.5918 12.8035 20.4593 11.9636C17.6508 12.6606 14.9992 13.4372 12.4356 14.2598C12.8479 17.4766 13.5448 21.1334 14.5118 24.7218C14.662 25.2792 14.8081 25.8248 14.9514 26.3594L14.9516 26.3603L14.9524 26.3634L14.9526 26.3639L14.973 26.4401C16.1833 25.9872 17.3746 25.5123 18.53 25.0259C20.1235 24.3552 21.6051 23.7165 22.9735 23.1135Z" fill="#141519"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.py

@@ -0,0 +1,15 @@
+from __future__ import annotations
+
+
+import typing
+
+from . import ppiochatcmpl
+
+
+class GiteeAIChatCompletions(ppiochatcmpl.PPIOChatCompletions):
+    """Gitee AI ChatCompletions API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://ai.gitee.com/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/giteeaichatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: gitee-ai-chat-completions
+  label:
+    en_US: Gitee AI
+    zh_Hans: Gitee AI
+  icon: giteeai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://ai.gitee.com/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./giteeaichatcmpl.py
+    attr: GiteeAIChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.py

@@ -0,0 +1,208 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+from .. import requester
+import openai.types.chat.chat_completion as chat_completion
+import re
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class JieKouAIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """接口 AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.jiekou.ai/openai',
+        'timeout': 120,
+    }
+
+    is_think: bool = False
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+        # print(chatcmpl_message.keys(), chatcmpl_message.values())
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        reasoning_content = chatcmpl_message['reasoning_content'] if 'reasoning_content' in chatcmpl_message else None
+
+        # deepseek的reasoner模型
+        chatcmpl_message['content'] = await self._process_thinking_content(
+            chatcmpl_message['content'], reasoning_content, remove_think
+        )
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            处理后的内容
+        """
+        if remove_think:
+            content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL)
+        else:
+            if reasoning_content is not None:
+                content = '<think>\n' + reasoning_content + '\n</think>\n' + content
+        return content
+
+    async def _make_msg_chunk(
+        self,
+        delta: dict[str, typing.Any],
+        idx: int,
+    ) -> provider_message.MessageChunk:
+        # 处理流式chunk和完整响应的差异
+        # print(chat_completion.choices[0])
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in delta or delta['role'] is None:
+            delta['role'] = 'assistant'
+
+        reasoning_content = delta['reasoning_content'] if 'reasoning_content' in delta else None
+
+        delta['content'] = '' if delta['content'] is None else delta['content']
+        # print(reasoning_content)
+
+        # deepseek的reasoner模型
+
+        if reasoning_content is not None:
+            delta['content'] += reasoning_content
+
+        message = provider_message.MessageChunk(**delta)
+
+        return message
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            # reasoning_content = delta.get('reasoning_content', '')
+
+            if remove_think:
+                if delta['content'] is not None:
+                    if '<think>' in delta['content'] and not thinking_started and not thinking_ended:
+                        thinking_started = True
+                        continue
+                    elif delta['content'] == r'</think>' and not thinking_ended:
+                        thinking_ended = True
+                        continue
+                    elif thinking_ended and delta['content'] == '\n\n' and thinking_started:
+                        thinking_started = False
+                        continue
+                    elif thinking_started and not thinking_ended:
+                        continue
+
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['id'] is None:
+                        tool_call['id'] = tool_id
+                    if tool_call['function']['name'] is None:
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['function']['arguments'] is None:
+                        tool_call['function']['arguments'] = ''
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1

src/langbot/pkg/provider/modelmgr/requesters/jiekouaichatcmpl.yaml

@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: jiekouai-chat-completions
+  label:
+    en_US: JieKou AI
+    zh_Hans: 接口 AI
+  icon: jiekouai.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.jiekou.ai/openai
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./jiekouaichatcmpl.py
+    attr: JieKouAIChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class LmStudioChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """LMStudio ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://127.0.0.1:1234/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/lmstudiochatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: lmstudio-chat-completions
+  label:
+    en_US: LM Studio
+    zh_Hans: LM Studio
+  icon: lmstudio.webp
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://127.0.0.1:1234/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: self-hosted
+execution:
+  python:
+    path: ./lmstudiochatcmpl.py
+    attr: LmStudioChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/modelscope.svg

@@ -0,0 +1 @@
+<svg id="_层_2" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 274.37 172.76"><defs><style>.cls-2{fill:#36cfd1}.cls-3{fill:#624aff}</style></defs><g id="_层_1-2"><path class="cls-3" d="M24.78 73.55h25.65V99.2H24.78zm99.14 25.66h25.65v25.65h-25.65zm76.95 25.65h-25.65v22.19h47.84V99.21h-22.19v25.65z"/><path class="cls-2" d="M149.57 73.55h25.65V99.2h-25.65zM24.78 47.9h25.65v25.65H24.78z"/><path class="cls-3" d="M223.06 73.55h25.65V99.2h-25.65z"/><path class="cls-2" d="M223.06 47.9h25.65v25.65h-25.65z"/><path class="cls-3" d="M175.22 25.71V47.9h25.65v25.65h22.19V25.71h-47.84z"/><path class="cls-2" d="M98.27 73.55h25.65V99.2H98.27z"/><path class="cls-3" d="M72.62 47.9h25.65V25.71H50.43v47.84h22.19V47.9zm0 51.31H50.43v47.84h47.84v-22.19H72.62V99.21z"/><path style="fill:none" d="M0 0h274.37v172.76H0z"/></g></svg>

src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.py

@@ -0,0 +1,389 @@
+from __future__ import annotations
+
+import asyncio
+import typing
+
+import openai
+import openai.types.chat.chat_completion as chat_completion
+import httpx
+
+from .. import entities, errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class ModelScopeChatCompletions(requester.ProviderAPIRequester):
+    """ModelScope ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api-inference.modelscope.cn/v1',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        self.client = openai.AsyncClient(
+            api_key='',
+            base_url=self.requester_cfg['base_url'],
+            timeout=self.requester_cfg['timeout'],
+            http_client=httpx.AsyncClient(trust_env=True, timeout=self.requester_cfg['timeout']),
+        )
+
+    async def _req(
+        self,
+        query: pipeline_query.Query,
+        args: dict,
+        extra_body: dict = {},
+        remove_think: bool = False,
+    ) -> list[dict[str, typing.Any]]:
+        args['stream'] = True
+
+        chunk = None
+
+        pending_content = ''
+
+        tool_calls = []
+
+        resp_gen: openai.AsyncStream = await self.client.chat.completions.create(**args, extra_body=extra_body)
+
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        tool_id = ''
+        tool_name = ''
+        message_delta = {}
+        async for chunk in resp_gen:
+            if not chunk or not chunk.id or not chunk.choices or not chunk.choices[0] or not chunk.choices[0].delta:
+                continue
+
+            delta = chunk.choices[0].delta.model_dump() if hasattr(chunk.choices[0], 'delta') else {}
+            reasoning_content = delta.get('reasoning_content')
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    pending_content += '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    pending_content += reasoning_content
+            elif thinking_started and not thinking_ended and delta.get('content'):
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                pending_content += '\n</think>\n' + delta.get('content')
+
+            if delta.get('content') is not None:
+                pending_content += delta.get('content')
+
+            if delta.get('tool_calls') is not None:
+                for tool_call in delta.get('tool_calls'):
+                    if tool_call['id'] != '':
+                        tool_id = tool_call['id']
+                    if tool_call['function']['name'] is not None:
+                        tool_name = tool_call['function']['name']
+                    if tool_call['function']['arguments'] is None:
+                        continue
+                    tool_call['id'] = tool_id
+                    tool_call['name'] = tool_name
+                    for tc in tool_calls:
+                        if tc['index'] == tool_call['index']:
+                            tc['function']['arguments'] += tool_call['function']['arguments']
+                            break
+                    else:
+                        tool_calls.append(tool_call)
+
+            if chunk.choices[0].finish_reason is not None:
+                break
+        message_delta['content'] = pending_content
+        message_delta['role'] = 'assistant'
+
+        message_delta['tool_calls'] = tool_calls if tool_calls else None
+        return [message_delta]
+
+    async def _make_msg(
+        self,
+        chat_completion: list[dict[str, typing.Any]],
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion[0]
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(query, args, extra_body=extra_args, remove_think=remove_think)
+
+        # 处理请求结果
+        message = await self._make_msg(resp)
+
+        return message
+
+    async def _req_stream(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        async for chunk in await self.client.chat.completions.create(**args, extra_body=extra_body):
+            yield chunk
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # 流式处理状态
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        # accumulated_reasoning = ''  # 仅用于判断何时结束思维链
+
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            reasoning_content = delta.get('reasoning_content', '')
+
+            # 处理 reasoning_content
+            if reasoning_content:
+                # accumulated_reasoning += reasoning_content
+                # 如果设置了 remove_think，跳过 reasoning_content
+                if remove_think:
+                    chunk_idx += 1
+                    continue
+
+                # 第一次出现 reasoning_content，添加 <think> 开始标签
+                if not thinking_started:
+                    thinking_started = True
+                    delta_content = '<think>\n' + reasoning_content
+                else:
+                    # 继续输出 reasoning_content
+                    delta_content = reasoning_content
+            elif thinking_started and not thinking_ended and delta_content:
+                # reasoning_content 结束，normal content 开始，添加 </think> 结束标签
+                thinking_ended = True
+                delta_content = '\n</think>\n' + delta_content
+
+            # 处理 content 中已有的 <think> 标签（如果需要移除）
+            # if delta_content and remove_think and '<think>' in delta_content:
+            #     import re
+            #
+            #     # 移除 <think> 标签及其内容
+            #     delta_content = re.sub(r'<think>.*?</think>', '', delta_content, flags=re.DOTALL)
+
+            # 处理工具调用增量
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] != '':
+                        tool_id = tool_call['id']
+                    if tool_call['function']['name'] is not None:
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+                    tool_call['id'] = tool_id
+                    tool_call['function']['name'] = tool_name
+                    tool_call['function']['arguments'] = (
+                        '' if tool_call['function']['arguments'] is None else tool_call['function']['arguments']
+                    )
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not reasoning_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1
+            # return
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: entities.LLMModelInfo,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            return await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')
+
+    async def invoke_llm_stream(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.MessageChunk:
+        req_messages = []  # req_messages 仅用于类内，外部同步由 query.messages 进行
+        for m in messages:
+            msg_dict = m.dict(exclude_none=True)
+            content = msg_dict.get('content')
+            if isinstance(content, list):
+                # 检查 content 列表中是否每个部分都是文本
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    # 将所有文本部分合并为一个字符串
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+
+        try:
+            async for item in self._closure_stream(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            ):
+                yield item
+
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+        except openai.BadRequestError as e:
+            if 'context_length_exceeded' in e.message:
+                raise errors.RequesterError(f'上文过长，请重置会话: {e.message}')
+            else:
+                raise errors.RequesterError(f'请求参数错误: {e.message}')
+        except openai.AuthenticationError as e:
+            raise errors.RequesterError(f'无效的 api-key: {e.message}')
+        except openai.NotFoundError as e:
+            raise errors.RequesterError(f'请求路径错误: {e.message}')
+        except openai.RateLimitError as e:
+            raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}')
+        except openai.APIError as e:
+            raise errors.RequesterError(f'请求错误: {e.message}')

src/langbot/pkg/provider/modelmgr/requesters/modelscopechatcmpl.yaml

@@ -0,0 +1,38 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: modelscope-chat-completions
+  label:
+    en_US: ModelScope
+    zh_Hans: 魔搭社区
+  icon: modelscope.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api-inference.modelscope.cn/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./modelscopechatcmpl.py
+    attr: ModelScopeChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.py

@@ -0,0 +1,60 @@
+from __future__ import annotations
+
+import typing
+
+
+from . import chatcmpl
+from .. import requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class MoonshotChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """Moonshot ChatCompletion API 请求器"""
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.moonshot.cn/v1',
+        'timeout': 120,
+    }
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages
+
+        # deepseek 不支持多模态，把content都转换成纯文字
+        for m in messages:
+            if 'content' in m and isinstance(m['content'], list):
+                m['content'] = ' '.join([c['text'] for c in m['content']])
+
+        # 删除空的，不知道干嘛的，直接删了。
+        # messages = [m for m in messages if m["content"].strip() != "" and ('tool_calls' not in m or not m['tool_calls'])]
+
+        args['messages'] = messages
+
+        # 发送请求
+        resp = await self._req(args, extra_body=extra_args)
+
+        # 处理请求结果
+        message = await self._make_msg(resp, remove_think)
+
+        return message

src/langbot/pkg/provider/modelmgr/requesters/moonshotchatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: moonshot-chat-completions
+  label:
+    en_US: Moonshot
+    zh_Hans: 月之暗面
+  icon: moonshot.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.moonshot.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./moonshotchatcmpl.py
+    attr: MoonshotChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class NewAPIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """New API ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://localhost:3000/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/newapichatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: new-api-chat-completions
+  label:
+    en_US: New API
+    zh_Hans: New API
+  icon: newapi.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://localhost:3000/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./newapichatcmpl.py
+    attr: NewAPIChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/ollamachat.py

@@ -0,0 +1,148 @@
+from __future__ import annotations
+
+import asyncio
+import os
+import typing
+from typing import Union, Mapping, Any, AsyncIterator
+import uuid
+import json
+
+import ollama
+
+from .. import errors, requester
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+REQUESTER_NAME: str = 'ollama-chat'
+
+
+class OllamaChatCompletions(requester.ProviderAPIRequester):
+    """Ollama平台 ChatCompletion API请求器"""
+
+    client: ollama.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'http://127.0.0.1:11434',
+        'timeout': 120,
+    }
+
+    async def initialize(self):
+        os.environ['OLLAMA_HOST'] = self.requester_cfg['base_url']
+        self.client = ollama.AsyncClient(timeout=self.requester_cfg['timeout'])
+
+    async def _req(
+        self,
+        args: dict,
+    ) -> Union[Mapping[str, Any], AsyncIterator[Mapping[str, Any]]]:
+        return await self.client.chat(**args)
+
+    async def _closure(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        args = extra_args.copy()
+        args['model'] = use_model.model_entity.name
+
+        messages: list[dict] = req_messages.copy()
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                text_content: list = []
+                image_urls: list = []
+                for me in msg['content']:
+                    if me['type'] == 'text':
+                        text_content.append(me['text'])
+                    elif me['type'] == 'image_base64':
+                        image_urls.append(me['image_base64'])
+
+                msg['content'] = '\n'.join(text_content)
+                msg['images'] = [url.split(',')[1] for url in image_urls]
+            if 'tool_calls' in msg:  # LangBot 内部以 str 存储 tool_calls 的参数，这里需要转换为 dict
+                for tool_call in msg['tool_calls']:
+                    tool_call['function']['arguments'] = json.loads(tool_call['function']['arguments'])
+        args['messages'] = messages
+
+        args['tools'] = []
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+            if tools:
+                args['tools'] = tools
+
+        resp = await self._req(args)
+        message: provider_message.Message = await self._make_msg(resp)
+        return message
+
+    async def _make_msg(self, chat_completions: ollama.ChatResponse) -> provider_message.Message:
+        message: ollama.Message = chat_completions.message
+        if message is None:
+            raise ValueError("chat_completions must contain a 'message' field")
+
+        ret_msg: provider_message.Message = None
+
+        if message.content is not None:
+            ret_msg = provider_message.Message(role='assistant', content=message.content)
+        if message.tool_calls is not None and len(message.tool_calls) > 0:
+            tool_calls: list[provider_message.ToolCall] = []
+
+            for tool_call in message.tool_calls:
+                tool_calls.append(
+                    provider_message.ToolCall(
+                        id=uuid.uuid4().hex,
+                        type='function',
+                        function=provider_message.FunctionCall(
+                            name=tool_call.function.name,
+                            arguments=json.dumps(tool_call.function.arguments),
+                        ),
+                    )
+                )
+            ret_msg.tool_calls = tool_calls
+
+        return ret_msg
+
+    async def invoke_llm(
+        self,
+        query: pipeline_query.Query,
+        model: requester.RuntimeLLMModel,
+        messages: typing.List[provider_message.Message],
+        funcs: typing.List[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message:
+        req_messages: list = []
+        for m in messages:
+            msg_dict: dict = m.dict(exclude_none=True)
+            content: Any = msg_dict.get('content')
+            if isinstance(content, list):
+                if all(isinstance(part, dict) and part.get('type') == 'text' for part in content):
+                    msg_dict['content'] = '\n'.join(part['text'] for part in content)
+            req_messages.append(msg_dict)
+        try:
+            return await self._closure(
+                query=query,
+                req_messages=req_messages,
+                use_model=model,
+                use_funcs=funcs,
+                extra_args=extra_args,
+                remove_think=remove_think,
+            )
+        except asyncio.TimeoutError:
+            raise errors.RequesterError('请求超时')
+
+    async def invoke_embedding(
+        self,
+        model: requester.RuntimeEmbeddingModel,
+        input_text: list[str],
+        extra_args: dict[str, typing.Any] = {},
+    ) -> list[list[float]]:
+        return (
+            await self.client.embed(
+                model=model.model_entity.name,
+                input=input_text,
+                **extra_args,
+            )
+        ).embeddings

src/langbot/pkg/provider/modelmgr/requesters/ollamachat.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: ollama-chat
+  label:
+    en_US: Ollama
+    zh_Hans: Ollama
+  icon: ollama.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: http://127.0.0.1:11434
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: self-hosted
+execution:
+  python:
+    path: ./ollamachat.py
+    attr: OllamaChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/openrouter.svg

@@ -0,0 +1,10 @@
+<svg width="25" height="21" viewBox="0 0 25 21" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M1.05858 10.1738C1.76158 10.1738 4.47988 9.56715 5.88589 8.77041C7.2919 7.97367 7.2919 7.97367 10.1977 5.91152C13.8766 3.30069 16.4779 4.17486 20.7428 4.17486" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M11.4182 7.63145L11.3787 7.65951C8.50565 9.69845 8.42504 9.75566 6.92566 10.6053C5.98567 11.138 4.74704 11.5436 3.75151 11.8089C2.80313 12.0615 1.71203 12.2829 1.05858 12.2829V8.06483C1.05075 8.06483 1.05422 8.06445 1.06984 8.06276C1.11491 8.05788 1.26116 8.04203 1.52896 7.9926C1.84599 7.9341 2.24205 7.84582 2.6657 7.73296C3.55657 7.49564 4.3801 7.1996 4.84612 6.93552C4.88175 6.91533 4.91635 6.89573 4.95001 6.87666C6.15007 6.19693 6.15657 6.19325 8.97708 4.1916C12.5199 1.67735 15.5815 1.83587 18.5849 1.99138C19.3056 2.0287 20.0229 2.06584 20.7428 2.06584V6.28388C19.6102 6.28388 18.6583 6.24193 17.8263 6.20527C15.1245 6.08621 13.685 6.02278 11.4182 7.63145Z" fill="black"/>
+<path d="M24.8671 4.20087L17.6613 8.36117V0.0405881L24.8671 4.20087Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.6378 0L24.9139 4.20087L17.6378 8.40176V0ZM17.6847 0.0811762V8.32058L24.8202 4.20087L17.6847 0.0811762Z" fill="black"/>
+<path d="M0.917975 10.1764C1.62098 10.1764 4.33927 10.7831 5.74529 11.5799C7.1513 12.3766 7.1513 12.3766 10.0571 14.4388C13.736 17.0496 16.3373 16.1754 20.6022 16.1754" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M0.929234 12.2875C0.913615 12.2858 0.910145 12.2854 0.917975 12.2854V8.06741C1.57142 8.06741 2.66253 8.28878 3.61091 8.54142C4.60644 8.80663 5.84507 9.21231 6.78506 9.74497C8.28444 10.5946 8.36505 10.6518 11.2381 12.6908L11.2776 12.7188C13.5444 14.3275 14.9839 14.2641 17.6857 14.145C18.5177 14.1083 19.4696 14.0664 20.6022 14.0664V18.2844C19.8823 18.2844 19.165 18.3216 18.4443 18.3589C15.4409 18.5144 12.3793 18.6729 8.83648 16.1587C6.01597 14.157 6.00947 14.1533 4.80941 13.4736C4.77575 13.4545 4.74115 13.4349 4.70551 13.4148C4.2395 13.1507 3.41597 12.8546 2.5251 12.6173C2.10145 12.5045 1.70538 12.4162 1.38836 12.3577C1.12056 12.3083 0.974309 12.2924 0.929234 12.2875Z" fill="black"/>
+<path d="M24.7265 16.1494L17.5207 11.9892V20.3097L24.7265 16.1494Z" fill="black"/>
+<path fill-rule="evenodd" clip-rule="evenodd" d="M17.4972 11.9486L24.7733 16.1494L17.4972 20.3503V11.9486ZM17.5441 12.0297V20.2691L24.6796 16.1494L17.5441 12.0297Z" fill="black"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import modelscopechatcmpl
+
+
+class OpenRouterChatCompletions(modelscopechatcmpl.ModelScopeChatCompletions):
+    """OpenRouter ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://openrouter.ai/api/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: openrouter-chat-completions
+  label:
+    en_US: OpenRouter
+    zh_Hans: OpenRouter
+  icon: openrouter.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://openrouter.ai/api/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./openrouterchatcmpl.py
+    attr: OpenRouterChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/ppio.svg

@@ -0,0 +1,3 @@
+<svg width="60" height="60" viewBox="0 0 60 60" fill="none" xmlns="http://www.w3.org/2000/svg">
+<path d="M29.7888 0.215881C13.3449 0.215881 0 13.5422 0 29.986C0 38.0916 3.24782 45.4527 8.51506 50.8223V30.0139C8.51506 24.3372 10.7299 18.9769 14.7408 14.966C18.7704 10.9365 24.112 8.74025 29.7981 8.74025H29.9749L29.7888 8.75886C41.5423 8.75886 51.0718 18.2883 51.0718 30.0326C51.0718 31.0562 50.9973 32.0613 50.8577 33.057L38.8343 20.9964C36.4333 18.5954 33.2134 17.2646 29.8074 17.2646C26.4013 17.2646 23.1907 18.5954 20.7805 20.9964C18.3609 23.4159 17.0394 26.6172 17.0394 30.0326C17.0394 33.4479 18.3702 36.6492 20.7805 39.0688C23.1814 41.4697 26.4013 42.8005 29.8074 42.8005C33.2134 42.8005 36.424 41.4697 38.8343 39.0688C41.077 36.826 42.3706 33.8946 42.5474 30.7584L49.6014 37.8403C46.4839 45.7319 38.797 51.3249 29.7981 51.3249C25.1357 51.3249 20.6874 49.8359 17.0301 47.072V56.9178C20.9014 58.7604 25.2195 59.7841 29.7794 59.7841C46.2233 59.7841 59.5682 46.4578 59.5682 30.0139C59.5868 13.5515 46.2512 0.225187 29.7981 0.225187L29.7888 0.215881Z" fill="#0062E2"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.py

@@ -0,0 +1,208 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+from .. import requester
+import openai.types.chat.chat_completion as chat_completion
+import re
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+class PPIOChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """欧派云 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.ppinfra.com/v3/openai',
+        'timeout': 120,
+    }
+
+    is_think: bool = False
+
+    async def _make_msg(
+        self,
+        chat_completion: chat_completion.ChatCompletion,
+        remove_think: bool,
+    ) -> provider_message.Message:
+        chatcmpl_message = chat_completion.choices[0].message.model_dump()
+        # print(chatcmpl_message.keys(), chatcmpl_message.values())
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None:
+            chatcmpl_message['role'] = 'assistant'
+
+        reasoning_content = chatcmpl_message['reasoning_content'] if 'reasoning_content' in chatcmpl_message else None
+
+        # deepseek的reasoner模型
+        chatcmpl_message['content'] = await self._process_thinking_content(
+            chatcmpl_message['content'], reasoning_content, remove_think
+        )
+
+        # 移除 reasoning_content 字段，避免传递给 Message
+        if 'reasoning_content' in chatcmpl_message:
+            del chatcmpl_message['reasoning_content']
+
+        message = provider_message.Message(**chatcmpl_message)
+
+        return message
+
+    async def _process_thinking_content(
+        self,
+        content: str,
+        reasoning_content: str = None,
+        remove_think: bool = False,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+            reasoning_content: reasoning_content 字段内容
+            remove_think: 是否移除思维链
+
+        Returns:
+            处理后的内容
+        """
+        if remove_think:
+            content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL)
+        else:
+            if reasoning_content is not None:
+                content = '<think>\n' + reasoning_content + '\n</think>\n' + content
+        return content
+
+    async def _make_msg_chunk(
+        self,
+        delta: dict[str, typing.Any],
+        idx: int,
+    ) -> provider_message.MessageChunk:
+        # 处理流式chunk和完整响应的差异
+        # print(chat_completion.choices[0])
+
+        # 确保 role 字段存在且不为 None
+        if 'role' not in delta or delta['role'] is None:
+            delta['role'] = 'assistant'
+
+        reasoning_content = delta['reasoning_content'] if 'reasoning_content' in delta else None
+
+        delta['content'] = '' if delta['content'] is None else delta['content']
+        # print(reasoning_content)
+
+        # deepseek的reasoner模型
+
+        if reasoning_content is not None:
+            delta['content'] += reasoning_content
+
+        message = provider_message.MessageChunk(**delta)
+
+        return message
+
+    async def _closure_stream(
+        self,
+        query: pipeline_query.Query,
+        req_messages: list[dict],
+        use_model: requester.RuntimeLLMModel,
+        use_funcs: list[resource_tool.LLMTool] = None,
+        extra_args: dict[str, typing.Any] = {},
+        remove_think: bool = False,
+    ) -> provider_message.Message | typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        self.client.api_key = use_model.token_mgr.get_token()
+
+        args = {}
+        args['model'] = use_model.model_entity.name
+
+        if use_funcs:
+            tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs)
+
+            if tools:
+                args['tools'] = tools
+
+        # 设置此次请求中的messages
+        messages = req_messages.copy()
+
+        # 检查vision
+        for msg in messages:
+            if 'content' in msg and isinstance(msg['content'], list):
+                for me in msg['content']:
+                    if me['type'] == 'image_base64':
+                        me['image_url'] = {'url': me['image_base64']}
+                        me['type'] = 'image_url'
+                        del me['image_base64']
+
+        args['messages'] = messages
+        args['stream'] = True
+
+        # tool_calls_map: dict[str, provider_message.ToolCall] = {}
+        chunk_idx = 0
+        thinking_started = False
+        thinking_ended = False
+        role = 'assistant'  # 默认角色
+        async for chunk in self._req_stream(args, extra_body=extra_args):
+            # 解析 chunk 数据
+            if hasattr(chunk, 'choices') and chunk.choices:
+                choice = chunk.choices[0]
+                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                finish_reason = getattr(choice, 'finish_reason', None)
+            else:
+                delta = {}
+                finish_reason = None
+
+            # 从第一个 chunk 获取 role，后续使用这个 role
+            if 'role' in delta and delta['role']:
+                role = delta['role']
+
+            # 获取增量内容
+            delta_content = delta.get('content', '')
+            # reasoning_content = delta.get('reasoning_content', '')
+
+            if remove_think:
+                if delta['content'] is not None:
+                    if '<think>' in delta['content'] and not thinking_started and not thinking_ended:
+                        thinking_started = True
+                        continue
+                    elif delta['content'] == r'</think>' and not thinking_ended:
+                        thinking_ended = True
+                        continue
+                    elif thinking_ended and delta['content'] == '\n\n' and thinking_started:
+                        thinking_started = False
+                        continue
+                    elif thinking_started and not thinking_ended:
+                        continue
+
+            # delta_tool_calls = None
+            if delta.get('tool_calls'):
+                for tool_call in delta['tool_calls']:
+                    if tool_call['id'] and tool_call['function']['name']:
+                        tool_id = tool_call['id']
+                        tool_name = tool_call['function']['name']
+
+                    if tool_call['id'] is None:
+                        tool_call['id'] = tool_id
+                    if tool_call['function']['name'] is None:
+                        tool_call['function']['name'] = tool_name
+                    if tool_call['function']['arguments'] is None:
+                        tool_call['function']['arguments'] = ''
+                    if tool_call['type'] is None:
+                        tool_call['type'] = 'function'
+
+            # 跳过空的第一个 chunk（只有 role 没有内容）
+            if chunk_idx == 0 and not delta_content and not delta.get('tool_calls'):
+                chunk_idx += 1
+                continue
+
+            # 构建 MessageChunk - 只包含增量内容
+            chunk_data = {
+                'role': role,
+                'content': delta_content if delta_content else None,
+                'tool_calls': delta.get('tool_calls'),
+                'is_final': bool(finish_reason),
+            }
+
+            # 移除 None 值
+            chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+
+            yield provider_message.MessageChunk(**chunk_data)
+            chunk_idx += 1

src/langbot/pkg/provider/modelmgr/requesters/ppiochatcmpl.yaml

@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: ppio-chat-completions
+  label:
+    en_US: ppio
+    zh_Hans: 派欧云
+  icon: ppio.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.ppinfra.com/v3/openai
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./ppiochatcmpl.py
+    attr: PPIOChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+
+
+class QHAIGCChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """启航 AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.qhaigc.com/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/qhaigcchatcmpl.yaml

@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: qhaigc-chat-completions
+  label:
+    en_US: QH AI
+    zh_Hans: 启航 AI
+  icon: qhaigc.png
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.qhaigc.net/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./qhaigcchatcmpl.py
+    attr: QHAIGCChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.py

@@ -0,0 +1,32 @@
+from __future__ import annotations
+
+import openai
+import typing
+
+from . import chatcmpl
+import openai.types.chat.chat_completion as chat_completion
+
+
+class ShengSuanYunChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """胜算云(ModelSpot.AI) ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://router.shengsuanyun.com/api/v1',
+        'timeout': 120,
+    }
+
+    async def _req(
+        self,
+        args: dict,
+        extra_body: dict = {},
+    ) -> chat_completion.ChatCompletion:
+        return await self.client.chat.completions.create(
+            **args,
+            extra_body=extra_body,
+            extra_headers={
+                'HTTP-Referer': 'https://langbot.app',
+                'X-Title': 'LangBot',
+            },
+        )

src/langbot/pkg/provider/modelmgr/requesters/shengsuanyun.yaml

@@ -0,0 +1,39 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: shengsuanyun-chat-completions
+  label:
+    en_US: ShengSuanYun
+    zh_Hans: 胜算云
+  icon: shengsuanyun.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://router.shengsuanyun.com/api/v1
+  - name: args
+    label:
+      en_US: Args
+      zh_Hans: 附加参数
+    type: object
+    required: true
+    default: {}
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: int
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./shengsuanyun.py
+    attr: ShengSuanYunChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/siliconflow.svg

@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" fill="none" version="1.1" width="128" height="128" viewBox="0 0 128 128"><g><g style="opacity:0;"><rect x="0" y="0" width="128" height="128" rx="0" fill="#FFFFFF" fill-opacity="1"/></g><g><path d="M100.74,12L93.2335,12C69.21260000000001,12,55.3672,27.3468,55.3672,50.8672L55.3672,54.8988C52.6011,54.1056,49.7377,53.7031,46.8601,53.7031C29.816499999999998,53.7031,16,67.5196,16,84.5632C16,101.6069,29.816499999999998,115.423,46.8601,115.423C63.9037,115.423,77.72030000000001,101.6069,77.72030000000001,84.5632C77.72030000000001,82.4902,77.51140000000001,80.4223,77.0967,78.3911L77.2197,78.3911L100.74,78.3911C106.9654,78.3681,112,73.3151,112,67.08959999999999C112,60.8642,106.9654,55.8111,100.74,55.7882L100.7362,55.7882L100.6985,55.7879L100.6606,55.7882L77.2197,55.7882L77.2195,49.8663C77.2195,40.8584,83.7252,34.352900000000005,93.2335,34.352900000000005L100.5653,34.352900000000005L100.5733,34.352900000000005L100.5812,34.352900000000005L100.74,34.352900000000005L100.74,34.352900000000005C106.8469,34.2605,111.7497,29.284,111.7497,23.1764C111.7497,17.06889,106.8469,12.0923454,100.74,12L100.74,12ZM56.0347,84.5632C56.0347,79.4962,51.9271,75.3885,46.8601,75.3885C41.793099999999995,75.3885,37.6854,79.4962,37.6854,84.5632C37.6854,89.6303,41.793099999999995,93.7378,46.8601,93.7378C51.9271,93.7378,56.0347,89.6303,56.0347,84.5632Z" fill-rule="evenodd" fill="#8358F6" fill-opacity="1"/></g></g></svg>

src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class SiliconFlowChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """SiliconFlow ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.siliconflow.cn/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/siliconflowchatcmpl.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: siliconflow-chat-completions
+  label:
+    en_US: SiliconFlow
+    zh_Hans: 硅基流动
+  icon: siliconflow.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.siliconflow.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./siliconflowchatcmpl.py
+    attr: SiliconFlowChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/tokenpony.svg

@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" width="450" height="280" viewBox="0 0 450 280" class="cursor-pointer h-24 flex-shrink-0 w-149"><g fill="none" fill-rule="nonzero"><path fill="#0005DE" d="M97.705 6.742c58.844 0 90.962 34.353 90.962 98.341v21.843c-15.118-2.479-30.297-6.573-45.558-12.3v-9.543c0-35.97-15.564-56.281-45.404-56.281s-45.404 20.31-45.404 56.281v72.48c0 36.117 15.65 56.818 45.404 56.818 26.78 0 42.133-16.768 44.936-46.452q22.397 6.473 44.905 9.356c-6.15 51.52-37.492 79.155-89.841 79.155-58.678 0-90.963-34.72-90.963-98.878v-72.479c0-63.988 32.119-98.34 90.963-98.34m253.627 0c58.844 0 90.963 34.353 90.963 98.341v72.48c0 64.157-32.285 98.877-90.963 98.877-52.438 0-83.797-27.729-89.874-79.415 15-2.026 29.965-5.252 44.887-9.67 2.658 30.042 18.036 47.026 44.987 47.026 29.755 0 45.404-20.7 45.404-56.819v-72.479c0-35.97-15.564-56.281-45.404-56.281s-45.403 20.31-45.403 56.281v8.778c-15.262 5.868-30.44 10.104-45.559 12.725v-21.503c0-63.988 32.118-98.34 90.962-98.34m-164.37 140.026.57.09.831.127-.83-.128a234.5 234.5 0 0 0 35.979 2.79q18.408.002 36.858-2.928l1.401-.226a242 242 0 0 0 1.45-.244l-1.037.175q.729-.12 1.458-.247l-.421.072 1.26-.219-.84.147a244 244 0 0 0 2.8-.5l-.792.144q.648-.117 1.298-.239l-.506.094q.66-.122 1.322-.248l-.816.154q.759-.142 1.518-.289l-.702.135a247 247 0 0 0 5.364-1.084l-.463.098a250 250 0 0 0 3.928-.864l-.785.178 1.45-.33-.665.152q.597-.137 1.193-.276l-.528.123a253 253 0 0 0 3.685-.882l-.254.063q.683-.168 1.366-.34l-1.112.277q.809-.2 1.618-.405l-.506.128q.818-.206 1.634-.417l-1.128.289q.71-.18 1.419-.365l1.506-.397a259 259 0 0 0 1.804-.488l-.433.119a261 261 0 0 0 3.751-1.053l-.681.196a264 264 0 0 0 1.735-.502l-1.054.306q.636-.184 1.272-.37l-.218.064 1.238-.366-1.02.302a266 266 0 0 0 2.936-.882l-1.026.312q.71-.214 1.42-.433l-.394.121q.675-.207 1.35-.418l-.955.297q.8-.246 1.6-.499l-.645.202q.86-.269 1.72-.543l-1.076.341q.666-.21 1.33-.423l-.254.082q.833-.266 1.665-.539l-1.41.457q.874-.28 1.75-.568l-.34.111q.702-.229 1.403-.462l-1.063.351q.818-.269 1.634-.542l-.571.19a276 276 0 0 0 4.038-1.378l-.735.256q.657-.228 1.315-.46l-.58.204q16.86-5.903 33.78-14.256l-7.114-12.453 42.909 6.553-13.148 45.541-7.734-13.537q-23.832 11.94-47.755 19.504l-.199.063a298 298 0 0 1-11.65 3.412 288 288 0 0 1-10.39 2.603 280 280 0 0 1-11.677 2.431 273 273 0 0 1-11.643 1.903 263.5 263.5 0 0 1-36.858 2.599q-17.437 0-34.844-2.323l-.227-.03q-.635-.085-1.27-.174l1.497.204a268 268 0 0 1-13.673-2.182 275 275 0 0 1-12.817-2.697 282 282 0 0 1-11.859-3.057 291 291 0 0 1-7.21-2.123c-17.23-5.314-34.43-12.334-51.59-21.051l-8.258 14.455-13.148-45.541 42.909-6.553-6.594 11.544q18.421 9.24 36.776 15.572l1.316.45 1.373.462-.831-.278q.795.267 1.589.53l-.758-.252q.632.211 1.264.419l-.506-.167q.642.212 1.284.42l-.778-.253a271 271 0 0 0 3.914 1.251l-.227-.07a267 267 0 0 0 3.428 1.046l-.194-.058 1.315.389-1.121-.331q.864.256 1.73.508l-.609-.177q.826.241 1.651.478l-1.043-.3 1.307.375-.264-.075q.802.228 1.603.452l-1.34-.377q1.034.294 2.067.58l-.727-.203q.713.2 1.426.394l-.699-.192q.62.171 1.237.338l-.538-.146a259 259 0 0 0 3.977 1.051l-.66-.17q.683.177 1.367.35l-.707-.18q.687.175 1.373.348l-.666-.168q.738.186 1.475.368l-.809-.2q.716.179 1.43.353l-.621-.153a253 253 0 0 0 3.766.898l-.308-.07q.735.17 1.472.336l-1.164-.266q.747.173 1.496.34l-.332-.074q.845.19 1.69.374l-1.358-.3q.932.21 1.864.41l-.505-.11q.726.159 1.452.313l-.947-.203q.72.156 1.44.307l-.493-.104q.684.144 1.368.286l-.875-.182q.743.155 1.485.306l-.61-.124q.932.192 1.864.376l-1.254-.252q.904.184 1.809.361l-.555-.109q.752.15 1.504.293l-.95-.184q.69.135 1.377.265l-.427-.081q.784.15 1.569.295l-1.142-.214q.717.136 1.434.268l-.292-.054a244 244 0 0 0 3.808.673l-.68-.116 1.063.18-.383-.064q1.076.18 2.152.352z"></path></g></svg>

src/langbot/pkg/provider/modelmgr/requesters/tokenpony.yaml

@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: tokenpony-chat-completions
+  label:
+    en_US: TokenPony
+    zh_Hans: 小马算力
+  icon: tokenpony.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.tokenpony.cn/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  - text-embedding
+  provider_category: maas
+execution:
+  python:
+    path: ./tokenponychatcmpl.py
+    attr: TokenPonyChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/tokenponychatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class TokenPonyChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """TokenPony ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.tokenpony.cn/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/volcark.svg

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg viewBox="-0.006 0 24.6978 24.9156" xmlns="http://www.w3.org/2000/svg">
+  <path d="M20.511 15.3019L17.2442 28.1928C17.2362 28.2282 17.2364 28.2649 17.2447 28.3001C17.2531 28.3354 17.2694 28.3683 17.2923 28.3964C17.3153 28.4244 17.3444 28.4468 17.3773 28.4619C17.4103 28.477 17.4462 28.4844 17.4825 28.4835H24.0137C24.0499 28.4844 24.0859 28.477 24.1188 28.4619C24.1518 28.4468 24.1809 28.4244 24.2038 28.3964C24.2268 28.3683 24.2431 28.3354 24.2514 28.3001C24.2598 28.2649 24.26 28.2282 24.252 28.1928L20.9685 15.3019C20.9541 15.2524 20.924 15.209 20.8827 15.178C20.8415 15.1471 20.7913 15.1304 20.7397 15.1304C20.6882 15.1304 20.638 15.1471 20.5968 15.178C20.5555 15.209 20.5254 15.2524 20.511 15.3019V15.3019Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M2.53051 18.2228L-5.28338e-06 28.1924C-0.00799016 28.2277 -0.00780431 28.2644 0.000538111 28.2997C0.00888053 28.335 0.0251596 28.3679 0.0481365 28.3959C0.0711133 28.4239 0.100182 28.4464 0.133131 28.4615C0.166079 28.4766 0.202039 28.484 0.238273 28.4831H5.28025C5.31649 28.484 5.35245 28.4766 5.38539 28.4615C5.41834 28.4464 5.44741 28.4239 5.47039 28.3959C5.49336 28.3679 5.50964 28.335 5.51799 28.2997C5.52633 28.2644 5.52651 28.2277 5.51853 28.1924L2.98563 18.2228C2.97054 18.1742 2.94032 18.1318 2.89938 18.1016C2.85844 18.0714 2.80892 18.0552 2.75807 18.0552C2.70722 18.0552 2.6577 18.0714 2.61676 18.1016C2.57582 18.1318 2.5456 18.1742 2.53051 18.2228V18.2228Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M6.99344 9.96839L2.38275 28.1919C2.37498 28.2263 2.37494 28.262 2.38262 28.2964C2.3903 28.3308 2.40552 28.363 2.42717 28.3908C2.44882 28.4186 2.47637 28.4413 2.50783 28.4572C2.53929 28.473 2.57388 28.4817 2.60911 28.4826H11.8329C11.8691 28.4835 11.9051 28.4761 11.938 28.461C11.971 28.4459 12 28.4235 12.023 28.3955C12.046 28.3675 12.0623 28.3345 12.0706 28.2993C12.079 28.264 12.0791 28.2273 12.0712 28.1919L7.44855 9.96839C7.43347 9.91982 7.40325 9.87736 7.36231 9.8472C7.32136 9.81705 7.27185 9.80078 7.221 9.80078C7.17015 9.80078 7.12063 9.81705 7.07969 9.8472C7.03874 9.87736 7.00852 9.91982 6.99344 9.96839Z" fill="#006EFF" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M14.9472 4.17346C14.9321 4.1249 14.9019 4.08244 14.861 4.05228C14.82 4.02213 14.7705 4.00586 14.7197 4.00586C14.6688 4.00586 14.6193 4.02213 14.5784 4.05228C14.5374 4.08244 14.5072 4.1249 14.4921 4.17346L8.18963 28.192C8.18165 28.2273 8.18183 28.264 8.19017 28.2993C8.19852 28.3346 8.2148 28.3675 8.23777 28.3955C8.26075 28.4235 8.28982 28.446 8.32277 28.4611C8.35572 28.4762 8.39168 28.4835 8.42791 28.4827H21.0233C21.0596 28.4835 21.0955 28.4762 21.1285 28.4611C21.1614 28.446 21.1905 28.4235 21.2135 28.3955C21.2364 28.3675 21.2527 28.3346 21.2611 28.2993C21.2694 28.264 21.2696 28.2273 21.2616 28.192L14.9472 4.17346Z" fill="#006EFF" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+  <path d="M10.3175 12.6188L6.31915 28.1903C6.31074 28.2258 6.31061 28.2628 6.31875 28.2984C6.3269 28.3339 6.34311 28.3672 6.36614 28.3955C6.38916 28.4238 6.41839 28.4465 6.45155 28.4617C6.48472 28.4769 6.52094 28.4844 6.55743 28.4834H14.535C14.5715 28.4844 14.6077 28.4769 14.6409 28.4617C14.674 28.4465 14.7033 28.4238 14.7263 28.3955C14.7493 28.3672 14.7655 28.3339 14.7737 28.2984C14.7818 28.2628 14.7817 28.2258 14.7733 28.1903L10.7726 12.6188C10.7575 12.5702 10.7273 12.5278 10.6863 12.4976C10.6454 12.4674 10.5959 12.4512 10.545 12.4512C10.4942 12.4512 10.4447 12.4674 10.4037 12.4976C10.3628 12.5278 10.3326 12.5702 10.3175 12.6188Z" fill="#00E5E5" transform="matrix(1.0178890228271484, 0, 0, 1.0178890228271484, -1.952212187461555e-7, -4.077521402283104)"/>
+</svg>

src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class VolcArkChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """火山方舟大模型平台 ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://ark.cn-beijing.volces.com/api/v3',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/volcarkchatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: volcark-chat-completions
+  label:
+    en_US: Volc Engine Ark
+    zh_Hans: 火山方舟
+  icon: volcark.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://ark.cn-beijing.volces.com/api/v3
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: maas
+execution:
+  python:
+    path: ./volcarkchatcmpl.py
+    attr: VolcArkChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/xai.svg

@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" fill="currentColor" viewBox="0 0 24 24" aria-hidden="true" class="" focusable="false" style="fill:currentColor;height:28px;width:28px"><path d="m3.005 8.858 8.783 12.544h3.904L6.908 8.858zM6.905 15.825 3 21.402h3.907l1.951-2.788zM16.585 2l-6.75 9.64 1.953 2.79L20.492 2zM17.292 7.965v13.437h3.2V3.395z"></path></svg>

src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class XaiChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """xAI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://api.x.ai/v1',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/xaichatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: xai-chat-completions
+  label:
+    en_US: xAI
+    zh_Hans: xAI
+  icon: xai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://api.x.ai/v1
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./xaichatcmpl.py
+    attr: XaiChatCompletions

src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.py

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+import typing
+import openai
+
+from . import chatcmpl
+
+
+class ZhipuAIChatCompletions(chatcmpl.OpenAIChatCompletions):
+    """智谱AI ChatCompletion API 请求器"""
+
+    client: openai.AsyncClient
+
+    default_config: dict[str, typing.Any] = {
+        'base_url': 'https://open.bigmodel.cn/api/paas/v4',
+        'timeout': 120,
+    }

src/langbot/pkg/provider/modelmgr/requesters/zhipuaichatcmpl.yaml

@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: LLMAPIRequester
+metadata:
+  name: zhipuai-chat-completions
+  label:
+    en_US: ZhipuAI
+    zh_Hans: 智谱 AI
+  icon: zhipuai.svg
+spec:
+  config:
+  - name: base_url
+    label:
+      en_US: Base URL
+      zh_Hans: 基础 URL
+    type: string
+    required: true
+    default: https://open.bigmodel.cn/api/paas/v4
+  - name: timeout
+    label:
+      en_US: Timeout
+      zh_Hans: 超时时间
+    type: integer
+    required: true
+    default: 120
+  support_type:
+  - llm
+  provider_category: manufacturer
+execution:
+  python:
+    path: ./zhipuaichatcmpl.py
+    attr: ZhipuAIChatCompletions

src/langbot/pkg/provider/modelmgr/token.py

@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+import typing
+
+
+class TokenManager:
+    """鉴权 Token 管理器"""
+
+    name: str
+
+    tokens: list[str]
+
+    using_token_index: typing.Optional[int] = 0
+
+    def __init__(self, name: str, tokens: list[str]):
+        self.name = name
+        self.tokens = tokens
+        self.using_token_index = 0
+
+    def get_token(self) -> str:
+        return self.tokens[self.using_token_index]
+
+    def next_token(self):
+        self.using_token_index = (self.using_token_index + 1) % len(self.tokens)

src/langbot/pkg/provider/runner.py

@@ -0,0 +1,41 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from ..core import app
+
+
+preregistered_runners: list[typing.Type[RequestRunner]] = []
+
+
+def runner_class(name: str):
+    """注册一个请求运行器"""
+
+    def decorator(cls: typing.Type[RequestRunner]) -> typing.Type[RequestRunner]:
+        cls.name = name
+        preregistered_runners.append(cls)
+        return cls
+
+    return decorator
+
+
+class RequestRunner(abc.ABC):
+    """请求运行器"""
+
+    name: str = None
+
+    ap: app.Application
+
+    pipeline_config: dict
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+    @abc.abstractmethod
+    async def run(
+        self, query: core_entities.Query
+    ) -> typing.AsyncGenerator[llm_entities.Message | llm_entities.MessageChunk, None]:
+        """运行请求"""
+        pass

src/langbot/pkg/provider/runners/cozeapi.py

@@ -0,0 +1,288 @@
+from __future__ import annotations
+
+import typing
+import json
+import base64
+
+from langbot.pkg.provider import runner
+from langbot.pkg.core import app
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+from langbot.pkg.utils import image
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+from langbot.libs.coze_server_api.client import AsyncCozeAPIClient
+
+
+@runner.runner_class('coze-api')
+class CozeAPIRunner(runner.RequestRunner):
+    """Coze API 对话请求器"""
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        self.pipeline_config = pipeline_config
+        self.ap = ap
+        self.agent_token = pipeline_config['ai']['coze-api']['api-key']
+        self.bot_id = pipeline_config['ai']['coze-api'].get('bot-id')
+        self.chat_timeout = pipeline_config['ai']['coze-api'].get('timeout')
+        self.auto_save_history = pipeline_config['ai']['coze-api'].get('auto_save_history')
+        self.api_base = pipeline_config['ai']['coze-api'].get('api-base')
+
+        self.coze = AsyncCozeAPIClient(self.agent_token, self.api_base)
+
+    def _process_thinking_content(
+        self,
+        content: str,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+        Returns:
+            (处理后的内容, 提取的思维链内容)
+        """
+        remove_think = self.pipeline_config.get('output', {}).get('misc', {}).get('remove-think', False)
+        thinking_content = ''
+        # 从 content 中提取 <think> 标签内容
+        if content and '<think>' in content and '</think>' in content:
+            import re
+
+            think_pattern = r'<think>(.*?)</think>'
+            think_matches = re.findall(think_pattern, content, re.DOTALL)
+            if think_matches:
+                thinking_content = '\n'.join(think_matches)
+                # 移除 content 中的 <think> 标签
+                content = re.sub(think_pattern, '', content, flags=re.DOTALL).strip()
+
+        # 根据 remove_think 参数决定是否保留思维链
+        if remove_think:
+            return content, ''
+        else:
+            # 如果有思维链内容，将其以 <think> 格式添加到 content 开头
+            if thinking_content:
+                content = f'<think>\n{thinking_content}\n</think>\n{content}'.strip()
+            return content, thinking_content
+
+    async def _preprocess_user_message(self, query: pipeline_query.Query) -> list[dict]:
+        """预处理用户消息，转换为Coze消息格式
+
+        Returns:
+            list[dict]: Coze消息列表
+        """
+        messages = []
+
+        if isinstance(query.user_message.content, list):
+            # 多模态消息处理
+            content_parts = []
+
+            for ce in query.user_message.content:
+                if ce.type == 'text':
+                    content_parts.append({'type': 'text', 'text': ce.text})
+                elif ce.type == 'image_base64':
+                    image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+                    file_bytes = base64.b64decode(image_b64)
+                    file_id = await self._get_file_id(file_bytes)
+                    content_parts.append({'type': 'image', 'file_id': file_id})
+                elif ce.type == 'file':
+                    # 处理文件，上传到Coze
+                    file_id = await self._get_file_id(ce.file)
+                    content_parts.append({'type': 'file', 'file_id': file_id})
+
+            # 创建多模态消息
+            if content_parts:
+                messages.append(
+                    {
+                        'role': 'user',
+                        'content': json.dumps(content_parts),
+                        'content_type': 'object_string',
+                        'meta_data': None,
+                    }
+                )
+
+        elif isinstance(query.user_message.content, str):
+            # 纯文本消息
+            messages.append(
+                {'role': 'user', 'content': query.user_message.content, 'content_type': 'text', 'meta_data': None}
+            )
+
+        return messages
+
+    async def _get_file_id(self, file) -> str:
+        """上传文件到Coze服务
+        Args:
+            file: 文件
+        Returns:
+            str: 文件ID
+        """
+        file_id = await self.coze.upload(file=file)
+        return file_id
+
+    async def _chat_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """调用聊天助手（非流式）
+
+        注意：由于cozepy没有提供非流式API，这里使用流式API并在结束后一次性返回完整内容
+        """
+        user_id = f'{query.launcher_type.value}_{query.launcher_id}'
+
+        # 预处理用户消息
+        additional_messages = await self._preprocess_user_message(query)
+
+        # 获取会话ID
+        conversation_id = None
+
+        # 收集完整内容
+        full_content = ''
+        full_reasoning = ''
+
+        try:
+            # 调用Coze API流式接口
+            async for chunk in self.coze.chat_messages(
+                bot_id=self.bot_id,
+                user_id=user_id,
+                additional_messages=additional_messages,
+                conversation_id=conversation_id,
+                timeout=self.chat_timeout,
+                auto_save_history=self.auto_save_history,
+                stream=True,
+            ):
+                self.ap.logger.debug(f'coze-chat-stream: {chunk}')
+
+                event_type = chunk.get('event')
+                data = chunk.get('data', {})
+                # Removed debug print statement to avoid cluttering logs in production
+
+                if event_type == 'conversation.message.delta':
+                    # 收集内容
+                    if 'content' in data:
+                        full_content += data.get('content', '')
+
+                    # 收集推理内容（如果有）
+                    if 'reasoning_content' in data:
+                        full_reasoning += data.get('reasoning_content', '')
+
+                elif event_type.split('.')[-1] == 'done':  # 本地部署coze时，结束event不为done
+                    # 保存会话ID
+                    if 'conversation_id' in data:
+                        conversation_id = data.get('conversation_id')
+
+                elif event_type == 'error':
+                    # 处理错误
+                    error_msg = f'Coze API错误: {data.get("message", "未知错误")}'
+                    yield provider_message.Message(
+                        role='assistant',
+                        content=error_msg,
+                    )
+                    return
+
+            # 处理思维链内容
+            content, thinking_content = self._process_thinking_content(full_content)
+            if full_reasoning:
+                remove_think = self.pipeline_config.get('output', {}).get('misc', {}).get('remove-think', False)
+                if not remove_think:
+                    content = f'<think>\n{full_reasoning}\n</think>\n{content}'.strip()
+
+            # 一次性返回完整内容
+            yield provider_message.Message(
+                role='assistant',
+                content=content,
+            )
+
+            # 保存会话ID
+            if conversation_id and query.session.using_conversation:
+                query.session.using_conversation.uuid = conversation_id
+
+        except Exception as e:
+            self.ap.logger.error(f'Coze API错误: {str(e)}')
+            yield provider_message.Message(
+                role='assistant',
+                content=f'Coze API调用失败: {str(e)}',
+            )
+
+    async def _chat_messages_chunk(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        """调用聊天助手（流式）"""
+        user_id = f'{query.launcher_type.value}_{query.launcher_id}'
+
+        # 预处理用户消息
+        additional_messages = await self._preprocess_user_message(query)
+
+        # 获取会话ID
+        conversation_id = None
+
+        start_reasoning = False
+        stop_reasoning = False
+        message_idx = 1
+        is_final = False
+        full_content = ''
+        remove_think = self.pipeline_config.get('output', {}).get('misc', {}).get('remove-think', False)
+
+        try:
+            # 调用Coze API流式接口
+            async for chunk in self.coze.chat_messages(
+                bot_id=self.bot_id,
+                user_id=user_id,
+                additional_messages=additional_messages,
+                conversation_id=conversation_id,
+                timeout=self.chat_timeout,
+                auto_save_history=self.auto_save_history,
+                stream=True,
+            ):
+                self.ap.logger.debug(f'coze-chat-stream-chunk: {chunk}')
+
+                event_type = chunk.get('event')
+                data = chunk.get('data', {})
+                content = ''
+
+                if event_type == 'conversation.message.delta':
+                    message_idx += 1
+                    # 处理内容增量
+                    if 'reasoning_content' in data and not remove_think:
+                        reasoning_content = data.get('reasoning_content', '')
+                        if reasoning_content and not start_reasoning:
+                            content = '<think/>\n'
+                            start_reasoning = True
+                        content += reasoning_content
+
+                    if 'content' in data:
+                        if data.get('content', ''):
+                            content += data.get('content', '')
+                            if not stop_reasoning and start_reasoning:
+                                content = f'</think>\n{content}'
+                                stop_reasoning = True
+
+                elif event_type.split('.')[-1] == 'done':  # 本地部署coze时，结束event不为done
+                    # 保存会话ID
+                    if 'conversation_id' in data:
+                        conversation_id = data.get('conversation_id')
+                        if query.session.using_conversation:
+                            query.session.using_conversation.uuid = conversation_id
+                    is_final = True
+
+                elif event_type == 'error':
+                    # 处理错误
+                    error_msg = f'Coze API错误: {data.get("message", "未知错误")}'
+                    yield provider_message.MessageChunk(role='assistant', content=error_msg, finish_reason='error')
+                    return
+                full_content += content
+                if message_idx % 8 == 0 or is_final:
+                    if full_content:
+                        yield provider_message.MessageChunk(role='assistant', content=full_content, is_final=is_final)
+
+        except Exception as e:
+            self.ap.logger.error(f'Coze API流式调用错误: {str(e)}')
+            yield provider_message.MessageChunk(
+                role='assistant', content=f'Coze API流式调用失败: {str(e)}', finish_reason='error'
+            )
+
+    async def run(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """运行"""
+        msg_seq = 0
+        if await query.adapter.is_stream_output_supported():
+            async for msg in self._chat_messages_chunk(query):
+                if isinstance(msg, provider_message.MessageChunk):
+                    msg_seq += 1
+                    msg.msg_sequence = msg_seq
+                yield msg
+        else:
+            async for msg in self._chat_messages(query):
+                yield msg

src/langbot/pkg/provider/runners/dashscopeapi.py

@@ -0,0 +1,354 @@
+from __future__ import annotations
+
+import typing
+import re
+
+import dashscope
+
+from .. import runner
+from ...core import app
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class DashscopeAPIError(Exception):
+    """Dashscope API 请求失败"""
+
+    def __init__(self, message: str):
+        self.message = message
+        super().__init__(self.message)
+
+
+@runner.runner_class('dashscope-app-api')
+class DashScopeAPIRunner(runner.RequestRunner):
+    "阿里云百炼DashsscopeAPI对话请求器"
+
+    # 运行器内部使用的配置
+    app_type: str  # 应用类型
+    app_id: str  # 应用ID
+    api_key: str  # API Key
+    references_quote: (
+        str  # 引用资料提示（当展示回答来源功能开启时，这个变量会作为引用资料名前的提示，可在provider.json中配置）
+    )
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        """初始化"""
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+        valid_app_types = ['agent', 'workflow']
+        self.app_type = self.pipeline_config['ai']['dashscope-app-api']['app-type']
+        # 检查配置文件中使用的应用类型是否支持
+        if self.app_type not in valid_app_types:
+            raise DashscopeAPIError(f'不支持的 Dashscope 应用类型: {self.app_type}')
+
+        # 初始化Dashscope 参数配置
+        self.app_id = self.pipeline_config['ai']['dashscope-app-api']['app-id']
+        self.api_key = self.pipeline_config['ai']['dashscope-app-api']['api-key']
+        self.references_quote = self.pipeline_config['ai']['dashscope-app-api']['references_quote']
+
+    def _replace_references(self, text, references_dict):
+        """阿里云百炼平台的自定义应用支持资料引用，此函数可以将引用标签替换为参考资料"""
+
+        # 匹配 <ref>[index_id]</ref> 形式的字符串
+        pattern = re.compile(r'<ref>\[(.*?)\]</ref>')
+
+        def replacement(match):
+            # 获取引用编号
+            ref_key = match.group(1)
+            if ref_key in references_dict:
+                # 如果有对应的参考资料按照provider.json中的reference_quote返回提示，来自哪个参考资料文件
+                return f'({self.references_quote} {references_dict[ref_key]})'
+            else:
+                # 如果没有对应的参考资料，保留原样
+                return match.group(0)
+
+        # 使用 re.sub() 进行替换
+        return pattern.sub(replacement, text)
+
+    async def _preprocess_user_message(self, query: pipeline_query.Query) -> tuple[str, list[str]]:
+        """预处理用户消息，提取纯文本，阿里云提供的上传文件方法过于复杂，暂不支持上传文件（包括图片）"""
+        plain_text = ''
+        image_ids = []
+        if isinstance(query.user_message.content, list):
+            for ce in query.user_message.content:
+                if ce.type == 'text':
+                    plain_text += ce.text
+                # 暂时不支持上传图片，保留代码以便后续扩展
+                # elif ce.type == "image_base64":
+                #     image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+                #     file_bytes = base64.b64decode(image_b64)
+                #     file = ("img.png", file_bytes, f"image/{image_format}")
+                #     file_upload_resp = await self.dify_client.upload_file(
+                #         file,
+                #         f"{query.session.launcher_type.value}_{query.session.launcher_id}",
+                #     )
+                #     image_id = file_upload_resp["id"]
+                #     image_ids.append(image_id)
+        elif isinstance(query.user_message.content, str):
+            plain_text = query.user_message.content
+
+        return plain_text, image_ids
+
+    async def _agent_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """Dashscope 智能体对话请求"""
+
+        # 局部变量
+        chunk = None  # 流式传输的块
+        pending_content = ''  # 待处理的Agent输出内容
+        references_dict = {}  # 用于存储引用编号和对应的参考资料
+        plain_text = ''  # 用户输入的纯文本信息
+        image_ids = []  # 用户输入的图片ID列表 （暂不支持）
+
+        think_start = False
+        think_end = False
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+        has_thoughts = True  # 获取思考过程
+        remove_think = self.pipeline_config['output'].get('misc', '').get('remove-think')
+        if remove_think:
+            has_thoughts = False
+        # 发送对话请求
+        response = dashscope.Application.call(
+            api_key=self.api_key,  # 智能体应用的API Key
+            app_id=self.app_id,  # 智能体应用的ID
+            prompt=plain_text,  # 用户输入的文本信息
+            stream=True,  # 流式输出
+            incremental_output=True,  # 增量输出，使用流式输出需要开启增量输出
+            session_id=query.session.using_conversation.uuid,  # 会话ID用于，多轮对话
+            has_thoughts=has_thoughts,
+            # rag_options={                                     # 主要用于文件交互，暂不支持
+            #     "session_file_ids": ["FILE_ID1"],             # FILE_ID1 替换为实际的临时文件ID,逗号隔开多个
+            # }
+        )
+        idx_chunk = 0
+        try:
+            is_stream = await query.adapter.is_stream_output_supported()
+
+        except AttributeError:
+            is_stream = False
+        if is_stream:
+            for chunk in response:
+                if chunk.get('status_code') != 200:
+                    raise DashscopeAPIError(
+                        f'Dashscope API 请求失败: status_code={chunk.get("status_code")} message={chunk.get("message")} request_id={chunk.get("request_id")} '
+                    )
+                if not chunk:
+                    continue
+                idx_chunk += 1
+                # 获取流式传输的output
+                stream_output = chunk.get('output', {})
+                stream_think = stream_output.get('thoughts', [])
+                if stream_think[0].get('thought'):
+                    if not think_start:
+                        think_start = True
+                        pending_content += f'<think>\n{stream_think[0].get("thought")}'
+                    else:
+                        # 继续输出 reasoning_content
+                        pending_content += stream_think[0].get('thought')
+                elif stream_think[0].get('thought') == '' and not think_end:
+                    think_end = True
+                    pending_content += '\n</think>\n'
+                if stream_output.get('text') is not None:
+                    pending_content += stream_output.get('text')
+                # 是否是流式最后一个chunk
+                is_final = False if stream_output.get('finish_reason', False) == 'null' else True
+
+                # 获取模型传出的参考资料列表
+                references_dict_list = stream_output.get('doc_references', [])
+
+                # 从模型传出的参考资料信息中提取用于替换的字典
+                if references_dict_list is not None:
+                    for doc in references_dict_list:
+                        if doc.get('index_id') is not None:
+                            references_dict[doc.get('index_id')] = doc.get('doc_name')
+
+                    # 将参考资料替换到文本中
+                    pending_content = self._replace_references(pending_content, references_dict)
+
+                if idx_chunk % 8 == 0 or is_final:
+                    yield provider_message.MessageChunk(
+                        role='assistant',
+                        content=pending_content,
+                        is_final=is_final,
+                    )
+            # 保存当前会话的session_id用于下次对话的语境
+            query.session.using_conversation.uuid = stream_output.get('session_id')
+        else:
+            for chunk in response:
+                if chunk.get('status_code') != 200:
+                    raise DashscopeAPIError(
+                        f'Dashscope API 请求失败: status_code={chunk.get("status_code")} message={chunk.get("message")} request_id={chunk.get("request_id")} '
+                    )
+                if not chunk:
+                    continue
+                idx_chunk += 1
+                # 获取流式传输的output
+                stream_output = chunk.get('output', {})
+                stream_think = stream_output.get('thoughts', [])
+                if stream_think[0].get('thought'):
+                    if not think_start:
+                        think_start = True
+                        pending_content += f'<think>\n{stream_think[0].get("thought")}'
+                    else:
+                        # 继续输出 reasoning_content
+                        pending_content += stream_think[0].get('thought')
+                elif stream_think[0].get('thought') == '' and not think_end:
+                    think_end = True
+                    pending_content += '\n</think>\n'
+                if stream_output.get('text') is not None:
+                    pending_content += stream_output.get('text')
+
+            # 保存当前会话的session_id用于下次对话的语境
+            query.session.using_conversation.uuid = stream_output.get('session_id')
+
+            # 获取模型传出的参考资料列表
+            references_dict_list = stream_output.get('doc_references', [])
+
+            # 从模型传出的参考资料信息中提取用于替换的字典
+            if references_dict_list is not None:
+                for doc in references_dict_list:
+                    if doc.get('index_id') is not None:
+                        references_dict[doc.get('index_id')] = doc.get('doc_name')
+
+                # 将参考资料替换到文本中
+                pending_content = self._replace_references(pending_content, references_dict)
+
+            yield provider_message.Message(
+                role='assistant',
+                content=pending_content,
+            )
+
+    async def _workflow_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """Dashscope 工作流对话请求"""
+
+        # 局部变量
+        chunk = None  # 流式传输的块
+        pending_content = ''  # 待处理的Agent输出内容
+        references_dict = {}  # 用于存储引用编号和对应的参考资料
+        plain_text = ''  # 用户输入的纯文本信息
+        image_ids = []  # 用户输入的图片ID列表 （暂不支持）
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        biz_params = {}
+        biz_params.update(query.variables)
+
+        # 发送对话请求
+        response = dashscope.Application.call(
+            api_key=self.api_key,  # 智能体应用的API Key
+            app_id=self.app_id,  # 智能体应用的ID
+            prompt=plain_text,  # 用户输入的文本信息
+            stream=True,  # 流式输出
+            incremental_output=True,  # 增量输出，使用流式输出需要开启增量输出
+            session_id=query.session.using_conversation.uuid,  # 会话ID用于，多轮对话
+            biz_params=biz_params,  # 工作流应用的自定义输入参数传递
+            flow_stream_mode='message_format',  # 消息模式，输出/结束节点的流式结果
+            # rag_options={                                     # 主要用于文件交互，暂不支持
+            #     "session_file_ids": ["FILE_ID1"],             # FILE_ID1 替换为实际的临时文件ID,逗号隔开多个
+            # }
+        )
+
+        # 处理API返回的流式输出
+        try:
+            is_stream = await query.adapter.is_stream_output_supported()
+
+        except AttributeError:
+            is_stream = False
+        idx_chunk = 0
+        if is_stream:
+            for chunk in response:
+                if chunk.get('status_code') != 200:
+                    raise DashscopeAPIError(
+                        f'Dashscope API 请求失败: status_code={chunk.get("status_code")} message={chunk.get("message")} request_id={chunk.get("request_id")} '
+                    )
+                if not chunk:
+                    continue
+                idx_chunk += 1
+                # 获取流式传输的output
+                stream_output = chunk.get('output', {})
+                if stream_output.get('workflow_message') is not None:
+                    pending_content += stream_output.get('workflow_message').get('message').get('content')
+                # if stream_output.get('text') is not None:
+                #     pending_content += stream_output.get('text')
+
+                is_final = False if stream_output.get('finish_reason', False) == 'null' else True
+
+                # 获取模型传出的参考资料列表
+                references_dict_list = stream_output.get('doc_references', [])
+
+                # 从模型传出的参考资料信息中提取用于替换的字典
+                if references_dict_list is not None:
+                    for doc in references_dict_list:
+                        if doc.get('index_id') is not None:
+                            references_dict[doc.get('index_id')] = doc.get('doc_name')
+
+                    # 将参考资料替换到文本中
+                    pending_content = self._replace_references(pending_content, references_dict)
+                if idx_chunk % 8 == 0 or is_final:
+                    yield provider_message.MessageChunk(
+                        role='assistant',
+                        content=pending_content,
+                        is_final=is_final,
+                    )
+
+            # 保存当前会话的session_id用于下次对话的语境
+            query.session.using_conversation.uuid = stream_output.get('session_id')
+
+        else:
+            for chunk in response:
+                if chunk.get('status_code') != 200:
+                    raise DashscopeAPIError(
+                        f'Dashscope API 请求失败: status_code={chunk.get("status_code")} message={chunk.get("message")} request_id={chunk.get("request_id")} '
+                    )
+                if not chunk:
+                    continue
+
+                # 获取流式传输的output
+                stream_output = chunk.get('output', {})
+                if stream_output.get('text') is not None:
+                    pending_content += stream_output.get('text')
+
+                is_final = False if stream_output.get('finish_reason', False) == 'null' else True
+
+            # 保存当前会话的session_id用于下次对话的语境
+            query.session.using_conversation.uuid = stream_output.get('session_id')
+
+            # 获取模型传出的参考资料列表
+            references_dict_list = stream_output.get('doc_references', [])
+
+            # 从模型传出的参考资料信息中提取用于替换的字典
+            if references_dict_list is not None:
+                for doc in references_dict_list:
+                    if doc.get('index_id') is not None:
+                        references_dict[doc.get('index_id')] = doc.get('doc_name')
+
+                # 将参考资料替换到文本中
+                pending_content = self._replace_references(pending_content, references_dict)
+
+            yield provider_message.Message(
+                role='assistant',
+                content=pending_content,
+            )
+
+    async def run(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """运行"""
+        msg_seq = 0
+        if self.app_type == 'agent':
+            async for msg in self._agent_messages(query):
+                if isinstance(msg, provider_message.MessageChunk):
+                    msg_seq += 1
+                    msg.msg_sequence = msg_seq
+                yield msg
+        elif self.app_type == 'workflow':
+            async for msg in self._workflow_messages(query):
+                if isinstance(msg, provider_message.MessageChunk):
+                    msg_seq += 1
+                    msg.msg_sequence = msg_seq
+                yield msg
+        else:
+            raise DashscopeAPIError(f'不支持的 Dashscope 应用类型: {self.app_type}')

src/langbot/pkg/provider/runners/difysvapi.py

@@ -0,0 +1,687 @@
+from __future__ import annotations
+
+import typing
+import json
+import uuid
+import base64
+
+
+from langbot.pkg.provider import runner
+from langbot.pkg.core import app
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+from langbot.pkg.utils import image
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+from langbot.libs.dify_service_api.v1 import client, errors
+
+
+@runner.runner_class('dify-service-api')
+class DifyServiceAPIRunner(runner.RequestRunner):
+    """Dify Service API 对话请求器"""
+
+    dify_client: client.AsyncDifyServiceClient
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+        valid_app_types = ['chat', 'agent', 'workflow']
+        if self.pipeline_config['ai']['dify-service-api']['app-type'] not in valid_app_types:
+            raise errors.DifyAPIError(
+                f'不支持的 Dify 应用类型: {self.pipeline_config["ai"]["dify-service-api"]["app-type"]}'
+            )
+
+        api_key = self.pipeline_config['ai']['dify-service-api']['api-key']
+
+        self.dify_client = client.AsyncDifyServiceClient(
+            api_key=api_key,
+            base_url=self.pipeline_config['ai']['dify-service-api']['base-url'],
+        )
+
+    def _process_thinking_content(
+        self,
+        content: str,
+    ) -> tuple[str, str]:
+        """处理思维链内容
+
+        Args:
+            content: 原始内容
+        Returns:
+            (处理后的内容, 提取的思维链内容)
+        """
+        remove_think = self.pipeline_config['output'].get('misc', '').get('remove-think')
+        thinking_content = ''
+        # 从 content 中提取 <think> 标签内容
+        if content and '<think>' in content and '</think>' in content:
+            import re
+
+            think_pattern = r'<think>(.*?)</think>'
+            think_matches = re.findall(think_pattern, content, re.DOTALL)
+            if think_matches:
+                thinking_content = '\n'.join(think_matches)
+                # 移除 content 中的 <think> 标签
+                content = re.sub(think_pattern, '', content, flags=re.DOTALL).strip()
+
+        # 3. 根据 remove_think 参数决定是否保留思维链
+        if remove_think:
+            return content, ''
+        else:
+            # 如果有思维链内容，将其以 <think> 格式添加到 content 开头
+            if thinking_content:
+                content = f'<think>\n{thinking_content}\n</think>\n{content}'.strip()
+            return content, thinking_content
+
+    async def _preprocess_user_message(self, query: pipeline_query.Query) -> tuple[str, list[str]]:
+        """预处理用户消息，提取纯文本，并将图片上传到 Dify 服务
+
+        Returns:
+            tuple[str, list[str]]: 纯文本和图片的 Dify 服务图片 ID
+        """
+        plain_text = ''
+        file_ids = []
+
+        if isinstance(query.user_message.content, list):
+            for ce in query.user_message.content:
+                if ce.type == 'text':
+                    plain_text += ce.text
+                elif ce.type == 'image_base64':
+                    image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+                    file_bytes = base64.b64decode(image_b64)
+                    file = ('img.png', file_bytes, f'image/{image_format}')
+                    file_upload_resp = await self.dify_client.upload_file(
+                        file,
+                        f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+                    )
+                    image_id = file_upload_resp['id']
+                    file_ids.append(image_id)
+                # elif ce.type == "file_url":
+                #     file_bytes = base64.b64decode(ce.file_url)
+                #     file_upload_resp = await self.dify_client.upload_file(
+                #         file_bytes,
+                #         f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+                #     )
+                #     file_id = file_upload_resp['id']
+                #     file_ids.append(file_id)
+        elif isinstance(query.user_message.content, str):
+            plain_text = query.user_message.content
+        # plain_text = "When the file content is readable, please read the content of this file. When the file is an image, describe the content of this image." if file_ids and not plain_text else plain_text
+        # plain_text = "The user message type cannot be parsed." if not file_ids and not plain_text else plain_text
+        # plain_text = plain_text if plain_text else "When the file content is readable, please read the content of this file. When the file is an image, describe the content of this image."
+        # print(self.pipeline_config['ai'])
+        plain_text = plain_text if plain_text else self.pipeline_config['ai']['dify-service-api']['base-prompt']
+
+        return plain_text, file_ids
+
+    async def _chat_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """调用聊天助手"""
+        cov_id = query.session.using_conversation.uuid or ''
+        query.variables['conversation_id'] = cov_id
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        mode = 'basic'  # 标记是基础编排还是工作流编排
+
+        basic_mode_pending_chunk = ''
+
+        inputs = {}
+
+        inputs.update(query.variables)
+
+        chunk = None  # 初始化chunk变量，防止在没有响应时引用错误
+
+        async for chunk in self.dify_client.chat_messages(
+            inputs=inputs,
+            query=plain_text,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            conversation_id=cov_id,
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-chat-chunk: ' + str(chunk))
+
+            if chunk['event'] == 'workflow_started':
+                mode = 'workflow'
+
+            if mode == 'workflow':
+                if chunk['event'] == 'node_finished':
+                    if chunk['data']['node_type'] == 'answer':
+                        content, _ = self._process_thinking_content(chunk['data']['outputs']['answer'])
+
+                        yield provider_message.Message(
+                            role='assistant',
+                            content=content,
+                        )
+            elif mode == 'basic':
+                if chunk['event'] == 'message':
+                    basic_mode_pending_chunk += chunk['answer']
+                elif chunk['event'] == 'message_end':
+                    content, _ = self._process_thinking_content(basic_mode_pending_chunk)
+                    yield provider_message.Message(
+                        role='assistant',
+                        content=content,
+                    )
+                    basic_mode_pending_chunk = ''
+
+        if chunk is None:
+            raise errors.DifyAPIError('Dify API 没有返回任何响应，请检查网络连接和API配置')
+
+        query.session.using_conversation.uuid = chunk['conversation_id']
+
+    async def _agent_chat_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """调用聊天助手"""
+        cov_id = query.session.using_conversation.uuid or ''
+        query.variables['conversation_id'] = cov_id
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'transfer_method': 'local_file',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        ignored_events = []
+
+        inputs = {}
+
+        inputs.update(query.variables)
+
+        pending_agent_message = ''
+
+        chunk = None  # 初始化chunk变量，防止在没有响应时引用错误
+
+        async for chunk in self.dify_client.chat_messages(
+            inputs=inputs,
+            query=plain_text,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            response_mode='streaming',
+            conversation_id=cov_id,
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-agent-chunk: ' + str(chunk))
+
+            if chunk['event'] in ignored_events:
+                continue
+
+            if chunk['event'] == 'agent_message' or chunk['event'] == 'message':
+                pending_agent_message += chunk['answer']
+            else:
+                if pending_agent_message.strip() != '':
+                    pending_agent_message = pending_agent_message.replace('</details>Action:', '</details>')
+                    content, _ = self._process_thinking_content(pending_agent_message)
+                    yield provider_message.Message(
+                        role='assistant',
+                        content=content,
+                    )
+                pending_agent_message = ''
+
+                if chunk['event'] == 'agent_thought':
+                    if chunk['tool'] != '' and chunk['observation'] != '':  # 工具调用结果，跳过
+                        continue
+
+                    if chunk['tool']:
+                        msg = provider_message.Message(
+                            role='assistant',
+                            tool_calls=[
+                                provider_message.ToolCall(
+                                    id=chunk['id'],
+                                    type='function',
+                                    function=provider_message.FunctionCall(
+                                        name=chunk['tool'],
+                                        arguments=json.dumps({}),
+                                    ),
+                                )
+                            ],
+                        )
+                        yield msg
+                if chunk['event'] == 'message_file':
+                    if chunk['type'] == 'image' and chunk['belongs_to'] == 'assistant':
+                        base_url = self.dify_client.base_url
+
+                        if base_url.endswith('/v1'):
+                            base_url = base_url[:-3]
+
+                        image_url = base_url + chunk['url']
+
+                        yield provider_message.Message(
+                            role='assistant',
+                            content=[provider_message.ContentElement.from_image_url(image_url)],
+                        )
+                if chunk['event'] == 'error':
+                    raise errors.DifyAPIError('dify 服务错误: ' + chunk['message'])
+
+        if chunk is None:
+            raise errors.DifyAPIError('Dify API 没有返回任何响应，请检查网络连接和API配置')
+
+        query.session.using_conversation.uuid = chunk['conversation_id']
+
+    async def _workflow_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """调用工作流"""
+
+        if not query.session.using_conversation.uuid:
+            query.session.using_conversation.uuid = str(uuid.uuid4())
+
+        query.variables['conversation_id'] = query.session.using_conversation.uuid
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'transfer_method': 'local_file',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        ignored_events = ['text_chunk', 'workflow_started']
+
+        inputs = {  # these variables are legacy variables, we need to keep them for compatibility
+            'langbot_user_message_text': plain_text,
+            'langbot_session_id': query.variables['session_id'],
+            'langbot_conversation_id': query.variables['conversation_id'],
+            'langbot_msg_create_time': query.variables['msg_create_time'],
+        }
+
+        inputs.update(query.variables)
+
+        async for chunk in self.dify_client.workflow_run(
+            inputs=inputs,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-workflow-chunk: ' + str(chunk))
+            if chunk['event'] in ignored_events:
+                continue
+
+            if chunk['event'] == 'node_started':
+                if chunk['data']['node_type'] == 'start' or chunk['data']['node_type'] == 'end':
+                    continue
+
+                msg = provider_message.Message(
+                    role='assistant',
+                    content=None,
+                    tool_calls=[
+                        provider_message.ToolCall(
+                            id=chunk['data']['node_id'],
+                            type='function',
+                            function=provider_message.FunctionCall(
+                                name=chunk['data']['title'],
+                                arguments=json.dumps({}),
+                            ),
+                        )
+                    ],
+                )
+
+                yield msg
+
+            elif chunk['event'] == 'workflow_finished':
+                if chunk['data']['error']:
+                    raise errors.DifyAPIError(chunk['data']['error'])
+                content, _ = self._process_thinking_content(chunk['data']['outputs']['summary'])
+
+                msg = provider_message.Message(
+                    role='assistant',
+                    content=content,
+                )
+
+                yield msg
+
+    async def _chat_messages_chunk(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        """调用聊天助手"""
+        cov_id = query.session.using_conversation.uuid or ''
+        query.variables['conversation_id'] = cov_id
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'transfer_method': 'local_file',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        basic_mode_pending_chunk = ''
+
+        inputs = {}
+
+        inputs.update(query.variables)
+        message_idx = 0
+
+        chunk = None  # 初始化chunk变量，防止在没有响应时引用错误
+
+        is_final = False
+        think_start = False
+        think_end = False
+
+        remove_think = self.pipeline_config['output'].get('misc', '').get('remove-think')
+
+        async for chunk in self.dify_client.chat_messages(
+            inputs=inputs,
+            query=plain_text,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            conversation_id=cov_id,
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-chat-chunk: ' + str(chunk))
+
+            # if chunk['event'] == 'workflow_started':
+            #     mode = 'workflow'
+            # if mode == 'workflow':
+            # elif mode == 'basic':
+            # 因为都只是返回的 message也没有工具调用什么的，暂时不分类
+            if chunk['event'] == 'message':
+                message_idx += 1
+                if remove_think:
+                    if '<think>' in chunk['answer'] and not think_start:
+                        think_start = True
+                        continue
+                    if '</think>' in chunk['answer'] and not think_end:
+                        import re
+
+                        content = re.sub(r'^\n</think>', '', chunk['answer'])
+                        basic_mode_pending_chunk += content
+                        think_end = True
+                    elif think_end:
+                        basic_mode_pending_chunk += chunk['answer']
+                    if think_start:
+                        continue
+
+                else:
+                    basic_mode_pending_chunk += chunk['answer']
+
+            if chunk['event'] == 'message_end':
+                is_final = True
+
+            if is_final or message_idx % 8 == 0:
+                # content, _ = self._process_thinking_content(basic_mode_pending_chunk)
+                yield provider_message.MessageChunk(
+                    role='assistant',
+                    content=basic_mode_pending_chunk,
+                    is_final=is_final,
+                )
+
+        if chunk is None:
+            raise errors.DifyAPIError('Dify API 没有返回任何响应，请检查网络连接和API配置')
+
+        query.session.using_conversation.uuid = chunk['conversation_id']
+
+    async def _agent_chat_messages_chunk(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        """调用聊天助手"""
+        cov_id = query.session.using_conversation.uuid or ''
+        query.variables['conversation_id'] = cov_id
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'transfer_method': 'local_file',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        ignored_events = []
+
+        inputs = {}
+
+        inputs.update(query.variables)
+
+        pending_agent_message = ''
+
+        chunk = None  # 初始化chunk变量，防止在没有响应时引用错误
+        message_idx = 0
+        is_final = False
+        think_start = False
+        think_end = False
+
+        remove_think = self.pipeline_config['output'].get('misc', '').get('remove-think')
+
+        async for chunk in self.dify_client.chat_messages(
+            inputs=inputs,
+            query=plain_text,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            response_mode='streaming',
+            conversation_id=cov_id,
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-agent-chunk: ' + str(chunk))
+
+            if chunk['event'] in ignored_events:
+                continue
+
+            if chunk['event'] == 'agent_message':
+                message_idx += 1
+                if remove_think:
+                    if '<think>' in chunk['answer'] and not think_start:
+                        think_start = True
+                        continue
+                    if '</think>' in chunk['answer'] and not think_end:
+                        import re
+
+                        content = re.sub(r'^\n</think>', '', chunk['answer'])
+                        pending_agent_message += content
+                        think_end = True
+                    elif think_end or not think_start:
+                        pending_agent_message += chunk['answer']
+                    if think_start:
+                        continue
+
+                else:
+                    pending_agent_message += chunk['answer']
+            elif chunk['event'] == 'message_end':
+                is_final = True
+            else:
+                if chunk['event'] == 'agent_thought':
+                    if chunk['tool'] != '' and chunk['observation'] != '':  # 工具调用结果，跳过
+                        continue
+                    message_idx += 1
+                    if chunk['tool']:
+                        msg = provider_message.MessageChunk(
+                            role='assistant',
+                            tool_calls=[
+                                provider_message.ToolCall(
+                                    id=chunk['id'],
+                                    type='function',
+                                    function=provider_message.FunctionCall(
+                                        name=chunk['tool'],
+                                        arguments=json.dumps({}),
+                                    ),
+                                )
+                            ],
+                        )
+                        yield msg
+                if chunk['event'] == 'message_file':
+                    message_idx += 1
+                    if chunk['type'] == 'image' and chunk['belongs_to'] == 'assistant':
+                        base_url = self.dify_client.base_url
+
+                        if base_url.endswith('/v1'):
+                            base_url = base_url[:-3]
+
+                        image_url = base_url + chunk['url']
+
+                        yield provider_message.MessageChunk(
+                            role='assistant',
+                            content=[provider_message.ContentElement.from_image_url(image_url)],
+                            is_final=is_final,
+                        )
+
+                if chunk['event'] == 'error':
+                    raise errors.DifyAPIError('dify 服务错误: ' + chunk['message'])
+            if message_idx % 8 == 0 or is_final:
+                yield provider_message.MessageChunk(
+                    role='assistant',
+                    content=pending_agent_message,
+                    is_final=is_final,
+                )
+
+        if chunk is None:
+            raise errors.DifyAPIError('Dify API 没有返回任何响应，请检查网络连接和API配置')
+
+        query.session.using_conversation.uuid = chunk['conversation_id']
+
+    async def _workflow_messages_chunk(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.MessageChunk, None]:
+        """调用工作流"""
+
+        if not query.session.using_conversation.uuid:
+            query.session.using_conversation.uuid = str(uuid.uuid4())
+
+        query.variables['conversation_id'] = query.session.using_conversation.uuid
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+
+        files = [
+            {
+                'type': 'image',
+                'transfer_method': 'local_file',
+                'upload_file_id': image_id,
+            }
+            for image_id in image_ids
+        ]
+
+        ignored_events = ['workflow_started']
+
+        inputs = {  # these variables are legacy variables, we need to keep them for compatibility
+            'langbot_user_message_text': plain_text,
+            'langbot_session_id': query.variables['session_id'],
+            'langbot_conversation_id': query.variables['conversation_id'],
+            'langbot_msg_create_time': query.variables['msg_create_time'],
+        }
+
+        inputs.update(query.variables)
+        messsage_idx = 0
+        is_final = False
+        think_start = False
+        think_end = False
+        workflow_contents = ''
+
+        remove_think = self.pipeline_config['output'].get('misc', '').get('remove-think')
+        async for chunk in self.dify_client.workflow_run(
+            inputs=inputs,
+            user=f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            files=files,
+            timeout=120,
+        ):
+            self.ap.logger.debug('dify-workflow-chunk: ' + str(chunk))
+            if chunk['event'] in ignored_events:
+                continue
+            if chunk['event'] == 'workflow_finished':
+                is_final = True
+                if chunk['data']['error']:
+                    raise errors.DifyAPIError(chunk['data']['error'])
+
+            if chunk['event'] == 'text_chunk':
+                messsage_idx += 1
+                if remove_think:
+                    if '<think>' in chunk['data']['text'] and not think_start:
+                        think_start = True
+                        continue
+                    if '</think>' in chunk['data']['text'] and not think_end:
+                        import re
+
+                        content = re.sub(r'^\n</think>', '', chunk['data']['text'])
+                        workflow_contents += content
+                        think_end = True
+                    elif think_end:
+                        workflow_contents += chunk['data']['text']
+                    if think_start:
+                        continue
+
+                else:
+                    workflow_contents += chunk['data']['text']
+
+            if chunk['event'] == 'node_started':
+                if chunk['data']['node_type'] == 'start' or chunk['data']['node_type'] == 'end':
+                    continue
+                messsage_idx += 1
+                msg = provider_message.MessageChunk(
+                    role='assistant',
+                    content=None,
+                    tool_calls=[
+                        provider_message.ToolCall(
+                            id=chunk['data']['node_id'],
+                            type='function',
+                            function=provider_message.FunctionCall(
+                                name=chunk['data']['title'],
+                                arguments=json.dumps({}),
+                            ),
+                        )
+                    ],
+                )
+
+                yield msg
+
+            if messsage_idx % 8 == 0 or is_final:
+                yield provider_message.MessageChunk(
+                    role='assistant',
+                    content=workflow_contents,
+                    is_final=is_final,
+                )
+
+    async def run(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """运行请求"""
+        if await query.adapter.is_stream_output_supported():
+            msg_idx = 0
+            if self.pipeline_config['ai']['dify-service-api']['app-type'] == 'chat':
+                async for msg in self._chat_messages_chunk(query):
+                    msg_idx += 1
+                    msg.msg_sequence = msg_idx
+                    yield msg
+            elif self.pipeline_config['ai']['dify-service-api']['app-type'] == 'agent':
+                async for msg in self._agent_chat_messages_chunk(query):
+                    msg_idx += 1
+                    msg.msg_sequence = msg_idx
+                    yield msg
+            elif self.pipeline_config['ai']['dify-service-api']['app-type'] == 'workflow':
+                async for msg in self._workflow_messages_chunk(query):
+                    msg_idx += 1
+                    msg.msg_sequence = msg_idx
+                    yield msg
+            else:
+                raise errors.DifyAPIError(
+                    f'不支持的 Dify 应用类型: {self.pipeline_config["ai"]["dify-service-api"]["app-type"]}'
+                )
+        else:
+            if self.pipeline_config['ai']['dify-service-api']['app-type'] == 'chat':
+                async for msg in self._chat_messages(query):
+                    yield msg
+            elif self.pipeline_config['ai']['dify-service-api']['app-type'] == 'agent':
+                async for msg in self._agent_chat_messages(query):
+                    yield msg
+            elif self.pipeline_config['ai']['dify-service-api']['app-type'] == 'workflow':
+                async for msg in self._workflow_messages(query):
+                    yield msg
+            else:
+                raise errors.DifyAPIError(
+                    f'不支持的 Dify 应用类型: {self.pipeline_config["ai"]["dify-service-api"]["app-type"]}'
+                )

src/langbot/pkg/provider/runners/langflowapi.py

@@ -0,0 +1,181 @@
+from __future__ import annotations
+
+import typing
+import json
+import httpx
+import uuid
+import traceback
+
+from .. import runner
+from ...core import app
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+@runner.runner_class('langflow-api')
+class LangflowAPIRunner(runner.RequestRunner):
+    """Langflow API 对话请求器"""
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+    async def _build_request_payload(self, query: pipeline_query.Query) -> dict:
+        """构建请求负载
+
+        Args:
+            query: 用户查询对象
+
+        Returns:
+            dict: 请求负载
+        """
+        # 获取用户消息文本
+        user_message_text = ''
+        if isinstance(query.user_message.content, str):
+            user_message_text = query.user_message.content
+        elif isinstance(query.user_message.content, list):
+            for item in query.user_message.content:
+                if item.type == 'text':
+                    user_message_text += item.text
+
+        # 从配置中获取 input_type 和 output_type，如果未配置则使用默认值
+        input_type = self.pipeline_config['ai']['langflow-api'].get('input_type', 'chat')
+        output_type = self.pipeline_config['ai']['langflow-api'].get('output_type', 'chat')
+
+        # 构建基本负载
+        payload = {
+            'output_type': output_type,
+            'input_type': input_type,
+            'input_value': user_message_text,
+            'session_id': str(uuid.uuid4()),
+        }
+
+        # 如果配置中有tweaks，则添加到负载中
+        tweaks = json.loads(self.pipeline_config['ai']['langflow-api'].get('tweaks'))
+        if tweaks:
+            payload['tweaks'] = tweaks
+
+        return payload
+
+    async def run(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message | provider_message.MessageChunk, None]:
+        """运行请求
+
+        Args:
+            query: 用户查询对象
+
+        Yields:
+            Message: 回复消息
+        """
+        # 检查是否支持流式输出
+        is_stream = False
+        try:
+            is_stream = await query.adapter.is_stream_output_supported()
+        except AttributeError:
+            is_stream = False
+
+        # 从配置中获取API参数
+        base_url = self.pipeline_config['ai']['langflow-api']['base-url']
+        api_key = self.pipeline_config['ai']['langflow-api']['api-key']
+        flow_id = self.pipeline_config['ai']['langflow-api']['flow-id']
+
+        # 构建API URL
+        url = f'{base_url.rstrip("/")}/api/v1/run/{flow_id}'
+
+        # 构建请求负载
+        payload = await self._build_request_payload(query)
+
+        # 设置请求头
+        headers = {'Content-Type': 'application/json', 'x-api-key': api_key}
+
+        # 发送请求
+        async with httpx.AsyncClient() as client:
+            if is_stream:
+                # 流式请求
+                async with client.stream('POST', url, json=payload, headers=headers, timeout=120.0) as response:
+                    print(response)
+                    response.raise_for_status()
+
+                    accumulated_content = ''
+                    message_count = 0
+
+                    async for line in response.aiter_lines():
+                        data_str = line
+
+                        if data_str.startswith('data: '):
+                            data_str = data_str[6:]  # 移除 "data: " 前缀
+
+                        try:
+                            data = json.loads(data_str)
+
+                            # 提取消息内容
+                            message_text = ''
+                            if 'outputs' in data and len(data['outputs']) > 0:
+                                output = data['outputs'][0]
+                                if 'outputs' in output and len(output['outputs']) > 0:
+                                    inner_output = output['outputs'][0]
+                                    if 'outputs' in inner_output and 'message' in inner_output['outputs']:
+                                        message_data = inner_output['outputs']['message']
+                                        if 'message' in message_data:
+                                            message_text = message_data['message']
+
+                            # 如果没有找到消息，尝试其他可能的路径
+                            if not message_text and 'messages' in data:
+                                messages = data['messages']
+                                if messages and len(messages) > 0:
+                                    message_text = messages[0].get('message', '')
+
+                            if message_text:
+                                # 更新累积内容
+                                accumulated_content = message_text
+                                message_count += 1
+
+                                # 每8条消息或有新内容时生成一个chunk
+                                if message_count % 8 == 0 or len(message_text) > 0:
+                                    yield provider_message.MessageChunk(
+                                        role='assistant', content=accumulated_content, is_final=False
+                                    )
+                        except json.JSONDecodeError:
+                            # 如果不是JSON，跳过这一行
+                            traceback.print_exc()
+                            continue
+
+                    # 发送最终消息
+                    yield provider_message.MessageChunk(role='assistant', content=accumulated_content, is_final=True)
+            else:
+                # 非流式请求
+                response = await client.post(url, json=payload, headers=headers, timeout=120.0)
+                response.raise_for_status()
+
+                # 解析响应
+                response_data = response.json()
+
+                # 提取消息内容
+                # 根据Langflow API文档，响应结构可能在outputs[0].outputs[0].outputs.message.message中
+                message_text = ''
+                if 'outputs' in response_data and len(response_data['outputs']) > 0:
+                    output = response_data['outputs'][0]
+                    if 'outputs' in output and len(output['outputs']) > 0:
+                        inner_output = output['outputs'][0]
+                        if 'outputs' in inner_output and 'message' in inner_output['outputs']:
+                            message_data = inner_output['outputs']['message']
+                            if 'message' in message_data:
+                                message_text = message_data['message']
+
+                # 如果没有找到消息，尝试其他可能的路径
+                if not message_text and 'messages' in response_data:
+                    messages = response_data['messages']
+                    if messages and len(messages) > 0:
+                        message_text = messages[0].get('message', '')
+
+                # 如果仍然没有找到消息，返回完整响应的字符串表示
+                if not message_text:
+                    message_text = json.dumps(response_data, ensure_ascii=False, indent=2)
+
+                # 生成回复消息
+                if is_stream:
+                    yield provider_message.MessageChunk(role='assistant', content=message_text, is_final=True)
+                else:
+                    reply_message = provider_message.Message(role='assistant', content=message_text)
+                    yield reply_message

src/langbot/pkg/provider/runners/localagent.py

@@ -0,0 +1,301 @@
+from __future__ import annotations
+
+import json
+import copy
+import typing
+from .. import runner
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+rag_combined_prompt_template = """
+The following are relevant context entries retrieved from the knowledge base. 
+Please use them to answer the user's message. 
+Respond in the same language as the user's input.
+
+<context>
+{rag_context}
+</context>
+
+<user_message>
+{user_message}
+</user_message>
+"""
+
+
+@runner.runner_class('local-agent')
+class LocalAgentRunner(runner.RequestRunner):
+    """本地Agent请求运行器"""
+
+    class ToolCallTracker:
+        """工具调用追踪器"""
+
+        def __init__(self):
+            self.active_calls: dict[str, dict] = {}
+            self.completed_calls: list[provider_message.ToolCall] = []
+
+    async def run(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message | provider_message.MessageChunk, None]:
+        """运行请求"""
+        pending_tool_calls = []
+
+        # Get knowledge bases list (new field)
+        kb_uuids = query.pipeline_config['ai']['local-agent'].get('knowledge-bases', [])
+
+        # Fallback to old field for backward compatibility
+        if not kb_uuids:
+            old_kb_uuid = query.pipeline_config['ai']['local-agent'].get('knowledge-base', '')
+            if old_kb_uuid and old_kb_uuid != '__none__':
+                kb_uuids = [old_kb_uuid]
+
+        user_message = copy.deepcopy(query.user_message)
+
+        user_message_text = ''
+
+        if isinstance(user_message.content, str):
+            user_message_text = user_message.content
+        elif isinstance(user_message.content, list):
+            for ce in user_message.content:
+                if ce.type == 'text':
+                    user_message_text += ce.text
+                    break
+
+        if kb_uuids and user_message_text:
+            # only support text for now
+            all_results = []
+
+            # Retrieve from each knowledge base
+            for kb_uuid in kb_uuids:
+                kb = await self.ap.rag_mgr.get_knowledge_base_by_uuid(kb_uuid)
+
+                if not kb:
+                    self.ap.logger.warning(f'Knowledge base {kb_uuid} not found, skipping')
+                    continue
+
+                result = await kb.retrieve(user_message_text, kb.knowledge_base_entity.top_k)
+
+                if result:
+                    all_results.extend(result)
+
+            final_user_message_text = ''
+
+            if all_results:
+                rag_context = '\n\n'.join(
+                    f'[{i + 1}] {entry.metadata.get("text", "")}' for i, entry in enumerate(all_results)
+                )
+                final_user_message_text = rag_combined_prompt_template.format(
+                    rag_context=rag_context, user_message=user_message_text
+                )
+
+            else:
+                final_user_message_text = user_message_text
+
+            self.ap.logger.debug(f'Final user message text: {final_user_message_text}')
+
+            for ce in user_message.content:
+                if ce.type == 'text':
+                    ce.text = final_user_message_text
+                    break
+
+        req_messages = query.prompt.messages.copy() + query.messages.copy() + [user_message]
+
+        try:
+            is_stream = await query.adapter.is_stream_output_supported()
+        except AttributeError:
+            is_stream = False
+
+        remove_think = query.pipeline_config['output'].get('misc', '').get('remove-think')
+
+        use_llm_model = await self.ap.model_mgr.get_model_by_uuid(query.use_llm_model_uuid)
+
+        if not is_stream:
+            # 非流式输出，直接请求
+
+            msg = await use_llm_model.requester.invoke_llm(
+                query,
+                use_llm_model,
+                req_messages,
+                query.use_funcs,
+                extra_args=use_llm_model.model_entity.extra_args,
+                remove_think=remove_think,
+            )
+            yield msg
+            final_msg = msg
+        else:
+            # 流式输出，需要处理工具调用
+            tool_calls_map: dict[str, provider_message.ToolCall] = {}
+            msg_idx = 0
+            accumulated_content = ''  # 从开始累积的所有内容
+            last_role = 'assistant'
+            msg_sequence = 1
+            async for msg in use_llm_model.requester.invoke_llm_stream(
+                query,
+                use_llm_model,
+                req_messages,
+                query.use_funcs,
+                extra_args=use_llm_model.model_entity.extra_args,
+                remove_think=remove_think,
+            ):
+                msg_idx = msg_idx + 1
+
+                # 记录角色
+                if msg.role:
+                    last_role = msg.role
+
+                # 累积内容
+                if msg.content:
+                    accumulated_content += msg.content
+
+                # 处理工具调用
+                if msg.tool_calls:
+                    for tool_call in msg.tool_calls:
+                        if tool_call.id not in tool_calls_map:
+                            tool_calls_map[tool_call.id] = provider_message.ToolCall(
+                                id=tool_call.id,
+                                type=tool_call.type,
+                                function=provider_message.FunctionCall(
+                                    name=tool_call.function.name if tool_call.function else '', arguments=''
+                                ),
+                            )
+                        if tool_call.function and tool_call.function.arguments:
+                            # 流式处理中，工具调用参数可能分多个chunk返回，需要追加而不是覆盖
+                            tool_calls_map[tool_call.id].function.arguments += tool_call.function.arguments
+                # continue
+                # 每8个chunk或最后一个chunk时，输出所有累积的内容
+                if msg_idx % 8 == 0 or msg.is_final:
+                    msg_sequence += 1
+                    yield provider_message.MessageChunk(
+                        role=last_role,
+                        content=accumulated_content,  # 输出所有累积内容
+                        tool_calls=list(tool_calls_map.values()) if (tool_calls_map and msg.is_final) else None,
+                        is_final=msg.is_final,
+                        msg_sequence=msg_sequence,
+                    )
+
+            # 创建最终消息用于后续处理
+            final_msg = provider_message.MessageChunk(
+                role=last_role,
+                content=accumulated_content,
+                tool_calls=list(tool_calls_map.values()) if tool_calls_map else None,
+                msg_sequence=msg_sequence,
+            )
+
+        pending_tool_calls = final_msg.tool_calls
+        first_content = final_msg.content
+        if isinstance(final_msg, provider_message.MessageChunk):
+            first_end_sequence = final_msg.msg_sequence
+
+        req_messages.append(final_msg)
+
+        # 持续请求，只要还有待处理的工具调用就继续处理调用
+        while pending_tool_calls:
+            for tool_call in pending_tool_calls:
+                try:
+                    func = tool_call.function
+
+                    parameters = json.loads(func.arguments)
+
+                    func_ret = await self.ap.tool_mgr.execute_func_call(func.name, parameters)
+                    if is_stream:
+                        msg = provider_message.MessageChunk(
+                            role='tool',
+                            content=json.dumps(func_ret, ensure_ascii=False),
+                            tool_call_id=tool_call.id,
+                        )
+                    else:
+                        msg = provider_message.Message(
+                            role='tool',
+                            content=json.dumps(func_ret, ensure_ascii=False),
+                            tool_call_id=tool_call.id,
+                        )
+
+                    yield msg
+
+                    req_messages.append(msg)
+                except Exception as e:
+                    # 工具调用出错，添加一个报错信息到 req_messages
+                    err_msg = provider_message.Message(role='tool', content=f'err: {e}', tool_call_id=tool_call.id)
+
+                    yield err_msg
+
+                    req_messages.append(err_msg)
+
+            if is_stream:
+                tool_calls_map = {}
+                msg_idx = 0
+                accumulated_content = ''  # 从开始累积的所有内容
+                last_role = 'assistant'
+                msg_sequence = first_end_sequence
+
+                async for msg in use_llm_model.requester.invoke_llm_stream(
+                    query,
+                    use_llm_model,
+                    req_messages,
+                    query.use_funcs,
+                    extra_args=use_llm_model.model_entity.extra_args,
+                    remove_think=remove_think,
+                ):
+                    msg_idx += 1
+
+                    # 记录角色
+                    if msg.role:
+                        last_role = msg.role
+
+                    # 第一次请求工具调用时的内容
+                    if msg_idx == 1:
+                        accumulated_content = first_content if first_content is not None else accumulated_content
+
+                    # 累积内容
+                    if msg.content:
+                        accumulated_content += msg.content
+
+                    # 处理工具调用
+                    if msg.tool_calls:
+                        for tool_call in msg.tool_calls:
+                            if tool_call.id not in tool_calls_map:
+                                tool_calls_map[tool_call.id] = provider_message.ToolCall(
+                                    id=tool_call.id,
+                                    type=tool_call.type,
+                                    function=provider_message.FunctionCall(
+                                        name=tool_call.function.name if tool_call.function else '', arguments=''
+                                    ),
+                                )
+                            if tool_call.function and tool_call.function.arguments:
+                                # 流式处理中，工具调用参数可能分多个chunk返回，需要追加而不是覆盖
+                                tool_calls_map[tool_call.id].function.arguments += tool_call.function.arguments
+
+                    # 每8个chunk或最后一个chunk时，输出所有累积的内容
+                    if msg_idx % 8 == 0 or msg.is_final:
+                        msg_sequence += 1
+                        yield provider_message.MessageChunk(
+                            role=last_role,
+                            content=accumulated_content,  # 输出所有累积内容
+                            tool_calls=list(tool_calls_map.values()) if (tool_calls_map and msg.is_final) else None,
+                            is_final=msg.is_final,
+                            msg_sequence=msg_sequence,
+                        )
+
+                final_msg = provider_message.MessageChunk(
+                    role=last_role,
+                    content=accumulated_content,
+                    tool_calls=list(tool_calls_map.values()) if tool_calls_map else None,
+                    msg_sequence=msg_sequence,
+                )
+            else:
+                # 处理完所有调用，再次请求
+                msg = await use_llm_model.requester.invoke_llm(
+                    query,
+                    use_llm_model,
+                    req_messages,
+                    query.use_funcs,
+                    extra_args=use_llm_model.model_entity.extra_args,
+                    remove_think=remove_think,
+                )
+
+                yield msg
+                final_msg = msg
+
+            pending_tool_calls = final_msg.tool_calls
+
+            req_messages.append(final_msg)

src/langbot/pkg/provider/runners/n8nsvapi.py

@@ -0,0 +1,160 @@
+from __future__ import annotations
+
+import typing
+import json
+import uuid
+import aiohttp
+
+from .. import runner
+from ...core import app
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class N8nAPIError(Exception):
+    """N8n API 请求失败"""
+
+    def __init__(self, message: str):
+        self.message = message
+        super().__init__(self.message)
+
+
+@runner.runner_class('n8n-service-api')
+class N8nServiceAPIRunner(runner.RequestRunner):
+    """N8n Service API 工作流请求器"""
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+        # 获取webhook URL
+        self.webhook_url = self.pipeline_config['ai']['n8n-service-api']['webhook-url']
+
+        # 获取超时设置，默认为120秒
+        self.timeout = self.pipeline_config['ai']['n8n-service-api'].get('timeout', 120)
+
+        # 获取输出键名，默认为response
+        self.output_key = self.pipeline_config['ai']['n8n-service-api'].get('output-key', 'response')
+
+        # 获取认证类型，默认为none
+        self.auth_type = self.pipeline_config['ai']['n8n-service-api'].get('auth-type', 'none')
+
+        # 根据认证类型获取相应的认证信息
+        if self.auth_type == 'basic':
+            self.basic_username = self.pipeline_config['ai']['n8n-service-api'].get('basic-username', '')
+            self.basic_password = self.pipeline_config['ai']['n8n-service-api'].get('basic-password', '')
+        elif self.auth_type == 'jwt':
+            self.jwt_secret = self.pipeline_config['ai']['n8n-service-api'].get('jwt-secret', '')
+            self.jwt_algorithm = self.pipeline_config['ai']['n8n-service-api'].get('jwt-algorithm', 'HS256')
+        elif self.auth_type == 'header':
+            self.header_name = self.pipeline_config['ai']['n8n-service-api'].get('header-name', '')
+            self.header_value = self.pipeline_config['ai']['n8n-service-api'].get('header-value', '')
+
+    async def _preprocess_user_message(self, query: pipeline_query.Query) -> str:
+        """预处理用户消息，提取纯文本
+
+        Returns:
+            str: 纯文本消息
+        """
+        plain_text = ''
+
+        if isinstance(query.user_message.content, list):
+            for ce in query.user_message.content:
+                if ce.type == 'text':
+                    plain_text += ce.text
+                # 注意：n8n webhook目前不支持直接处理图片，如需支持可在此扩展
+        elif isinstance(query.user_message.content, str):
+            plain_text = query.user_message.content
+
+        return plain_text
+
+    async def _call_webhook(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """调用n8n webhook"""
+        # 生成会话ID（如果不存在）
+        if not query.session.using_conversation.uuid:
+            query.session.using_conversation.uuid = str(uuid.uuid4())
+
+        # 预处理用户消息
+        plain_text = await self._preprocess_user_message(query)
+
+        # 准备请求数据
+        payload = {
+            # 基本消息内容
+            'message': plain_text,
+            'user_message_text': plain_text,
+            'conversation_id': query.session.using_conversation.uuid,
+            'session_id': query.variables.get('session_id', ''),
+            'user_id': f'{query.session.launcher_type.value}_{query.session.launcher_id}',
+            'msg_create_time': query.variables.get('msg_create_time', ''),
+        }
+
+        # 添加所有变量到payload
+        payload.update(query.variables)
+
+        try:
+            # 准备请求头和认证信息
+            headers = {}
+            auth = None
+
+            # 根据认证类型设置相应的认证信息
+            if self.auth_type == 'basic':
+                # 使用Basic认证
+                auth = aiohttp.BasicAuth(self.basic_username, self.basic_password)
+                self.ap.logger.debug(f'using basic auth: {self.basic_username}')
+            elif self.auth_type == 'jwt':
+                # 使用JWT认证
+                import jwt
+                import time
+
+                # 创建JWT令牌
+                payload_jwt = {
+                    'exp': int(time.time()) + 3600,  # 1小时过期
+                    'iat': int(time.time()),
+                    'sub': 'n8n-webhook',
+                }
+                token = jwt.encode(payload_jwt, self.jwt_secret, algorithm=self.jwt_algorithm)
+
+                # 添加到Authorization头
+                headers['Authorization'] = f'Bearer {token}'
+                self.ap.logger.debug('using jwt auth')
+            elif self.auth_type == 'header':
+                # 使用自定义请求头认证
+                headers[self.header_name] = self.header_value
+                self.ap.logger.debug(f'using header auth: {self.header_name}')
+            else:
+                self.ap.logger.debug('no auth')
+
+            # 调用webhook
+            async with aiohttp.ClientSession() as session:
+                async with session.post(
+                    self.webhook_url, json=payload, headers=headers, auth=auth, timeout=self.timeout
+                ) as response:
+                    if response.status != 200:
+                        error_text = await response.text()
+                        self.ap.logger.error(f'n8n webhook call failed: {response.status}, {error_text}')
+                        raise Exception(f'n8n webhook call failed: {response.status}, {error_text}')
+
+                    # 解析响应
+                    response_data = await response.json()
+                    self.ap.logger.debug(f'n8n webhook response: {response_data}')
+
+                    # 从响应中提取输出
+                    if self.output_key in response_data:
+                        output_content = response_data[self.output_key]
+                    else:
+                        # 如果没有指定的输出键，则使用整个响应
+                        output_content = json.dumps(response_data, ensure_ascii=False)
+
+                    # 返回消息
+                    yield provider_message.Message(
+                        role='assistant',
+                        content=output_content,
+                    )
+        except Exception as e:
+            self.ap.logger.error(f'n8n webhook call exception: {str(e)}')
+            raise N8nAPIError(f'n8n webhook call exception: {str(e)}')
+
+    async def run(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """运行请求"""
+        async for msg in self._call_webhook(query):
+            yield msg

src/langbot/pkg/provider/runners/tboxapi.py

@@ -0,0 +1,202 @@
+from __future__ import annotations
+
+import typing
+import json
+import base64
+import tempfile
+import os
+
+from tboxsdk.tbox import TboxClient
+from tboxsdk.model.file import File, FileType
+
+from .. import runner
+from ...core import app
+from ...utils import image
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+import langbot_plugin.api.entities.builtin.provider.message as provider_message
+
+
+class TboxAPIError(Exception):
+    """TBox API 请求失败"""
+
+    def __init__(self, message: str):
+        self.message = message
+        super().__init__(self.message)
+
+
+@runner.runner_class('tbox-app-api')
+class TboxAPIRunner(runner.RequestRunner):
+    "蚂蚁百宝箱API对话请求器"
+
+    # 运行器内部使用的配置
+    app_id: str  # 蚂蚁百宝箱平台中的应用ID
+    api_key: str  # 在蚂蚁百宝箱平台中申请的令牌
+
+    def __init__(self, ap: app.Application, pipeline_config: dict):
+        """初始化"""
+        self.ap = ap
+        self.pipeline_config = pipeline_config
+
+        # 初始化Tbox 参数配置
+        self.app_id = self.pipeline_config['ai']['tbox-app-api']['app-id']
+        self.api_key = self.pipeline_config['ai']['tbox-app-api']['api-key']
+
+        # 初始化Tbox client
+        self.tbox_client = TboxClient(authorization=self.api_key)
+
+    async def _preprocess_user_message(self, query: pipeline_query.Query) -> tuple[str, list[str]]:
+        """预处理用户消息，提取纯文本，并将图片上传到 Tbox 服务
+
+        Returns:
+            tuple[str, list[str]]: 纯文本和图片的 Tbox 文件ID
+        """
+        plain_text = ''
+        image_ids = []
+
+        if isinstance(query.user_message.content, list):
+            for ce in query.user_message.content:
+                if ce.type == 'text':
+                    plain_text += ce.text
+                elif ce.type == 'image_base64':
+                    image_b64, image_format = await image.extract_b64_and_format(ce.image_base64)
+                    # 创建临时文件
+                    file_bytes = base64.b64decode(image_b64)
+                    try:
+                        with tempfile.NamedTemporaryFile(suffix=f'.{image_format}', delete=False) as tmp_file:
+                            tmp_file.write(file_bytes)
+                            tmp_file_path = tmp_file.name
+                        file_upload_resp = self.tbox_client.upload_file(tmp_file_path)
+                        image_id = file_upload_resp.get('data', '')
+                        image_ids.append(image_id)
+                    finally:
+                        # 清理临时文件
+                        if os.path.exists(tmp_file_path):
+                            os.unlink(tmp_file_path)
+        elif isinstance(query.user_message.content, str):
+            plain_text = query.user_message.content
+
+        return plain_text, image_ids
+
+    async def _agent_messages(
+        self, query: pipeline_query.Query
+    ) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """TBox 智能体对话请求"""
+
+        plain_text, image_ids = await self._preprocess_user_message(query)
+        remove_think = self.pipeline_config['output'].get('misc', {}).get('remove-think')
+
+        try:
+            is_stream = await query.adapter.is_stream_output_supported()
+        except AttributeError:
+            is_stream = False
+
+        # 获取Tbox的conversation_id
+        conversation_id = query.session.using_conversation.uuid or None
+
+        files = None
+        if image_ids:
+            files = [File(file_id=image_id, type=FileType.IMAGE) for image_id in image_ids]
+
+        # 发送对话请求
+        response = self.tbox_client.chat(
+            app_id=self.app_id,  # Tbox中智能体应用的ID
+            user_id=query.bot_uuid,  # 用户ID
+            query=plain_text,  # 用户输入的文本信息
+            stream=is_stream,  # 是否流式输出
+            conversation_id=conversation_id,  # 会话ID，为None时Tbox会自动创建一个新会话
+            files=files,  # 图片内容
+        )
+
+        if is_stream:
+            # 解析Tbox流式输出内容，并发送给上游
+            for chunk in self._process_stream_message(response, query, remove_think):
+                yield chunk
+        else:
+            message = self._process_non_stream_message(response, query, remove_think)
+            yield provider_message.Message(
+                role='assistant',
+                content=message,
+            )
+
+    def _process_non_stream_message(self, response: typing.Dict, query: pipeline_query.Query, remove_think: bool):
+        if response.get('errorCode') != '0':
+            raise TboxAPIError(f'Tbox API 请求失败: {response.get("errorMsg", "")}')
+        payload = response.get('data', {})
+        conversation_id = payload.get('conversationId', '')
+        query.session.using_conversation.uuid = conversation_id
+        thinking_content = payload.get('reasoningContent', [])
+        result = ''
+        if thinking_content and not remove_think:
+            result += f'<think>\n{thinking_content[0].get("text", "")}\n</think>\n'
+        content = payload.get('result', [])
+        if content:
+            result += content[0].get('chunk', '')
+        return result
+
+    def _process_stream_message(
+        self, response: typing.Generator[dict], query: pipeline_query.Query, remove_think: bool
+    ):
+        idx_msg = 0
+        pending_content = ''
+        conversation_id = None
+        think_start = False
+        think_end = False
+        for chunk in response:
+            if chunk.get('type', '') == 'chunk':
+                """
+                Tbox返回的消息内容chunk结构
+                {'lane': 'default', 'payload': {'conversationId': '20250918tBI947065406', 'messageId': '20250918TB1f53230954', 'text': '️'}, 'type': 'chunk'}
+                """
+                # 如果包含思考过程，拼接</think>
+                if think_start and not think_end:
+                    pending_content += '\n</think>\n'
+                    think_end = True
+
+                payload = chunk.get('payload', {})
+                if not conversation_id:
+                    conversation_id = payload.get('conversationId')
+                    query.session.using_conversation.uuid = conversation_id
+                if payload.get('text'):
+                    idx_msg += 1
+                    pending_content += payload.get('text')
+            elif chunk.get('type', '') == 'thinking' and not remove_think:
+                """
+                Tbox返回的思考过程chunk结构
+                {'payload': '{"ext_data":{"text":"日期"},"event":"flow.node.llm.thinking","entity":{"node_type":"text-completion","execute_id":"6","group_id":0,"parent_execute_id":"6","node_name":"模型推理","node_id":"TC_5u6gl0"}}', 'type': 'thinking'}
+                """
+                payload = json.loads(chunk.get('payload', '{}'))
+                if payload.get('ext_data', {}).get('text'):
+                    idx_msg += 1
+                    content = payload.get('ext_data', {}).get('text')
+                    if not think_start:
+                        think_start = True
+                        pending_content += f'<think>\n{content}'
+                    else:
+                        pending_content += content
+            elif chunk.get('type', '') == 'error':
+                raise TboxAPIError(
+                    f'Tbox API 请求失败: status_code={chunk.get("status_code")} message={chunk.get("message")} request_id={chunk.get("request_id")} '
+                )
+
+            if idx_msg % 8 == 0:
+                yield provider_message.MessageChunk(
+                    role='assistant',
+                    content=pending_content,
+                    is_final=False,
+                )
+
+        # Tbox不返回END事件，默认发一个最终消息
+        yield provider_message.MessageChunk(
+            role='assistant',
+            content=pending_content,
+            is_final=True,
+        )
+
+    async def run(self, query: pipeline_query.Query) -> typing.AsyncGenerator[provider_message.Message, None]:
+        """运行"""
+        msg_seq = 0
+        async for msg in self._agent_messages(query):
+            if isinstance(msg, provider_message.MessageChunk):
+                msg_seq += 1
+                msg.msg_sequence = msg_seq
+            yield msg

src/langbot/pkg/provider/session/sessionmgr.py

@@ -0,0 +1,75 @@
+from __future__ import annotations
+
+import asyncio
+
+from ...core import app
+from langbot_plugin.api.entities.builtin.provider import message as provider_message, prompt as provider_prompt
+import langbot_plugin.api.entities.builtin.provider.session as provider_session
+import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
+
+
+class SessionManager:
+    """会话管理器"""
+
+    ap: app.Application
+
+    session_list: list[provider_session.Session]
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+        self.session_list = []
+
+    async def initialize(self):
+        pass
+
+    async def get_session(self, query: pipeline_query.Query) -> provider_session.Session:
+        """获取会话"""
+        for session in self.session_list:
+            if query.launcher_type == session.launcher_type and query.launcher_id == session.launcher_id:
+                return session
+
+        session_concurrency = self.ap.instance_config.data['concurrency']['session']
+
+        session = provider_session.Session(
+            launcher_type=query.launcher_type,
+            launcher_id=query.launcher_id,
+        )
+        session._semaphore = asyncio.Semaphore(session_concurrency)
+        self.session_list.append(session)
+        return session
+
+    async def get_conversation(
+        self,
+        query: pipeline_query.Query,
+        session: provider_session.Session,
+        prompt_config: list[dict],
+        pipeline_uuid: str,
+        bot_uuid: str,
+    ) -> provider_session.Conversation:
+        """获取对话或创建对话"""
+
+        if not session.conversations:
+            session.conversations = []
+
+        # set prompt
+        prompt_messages = []
+
+        for prompt_message in prompt_config:
+            prompt_messages.append(provider_message.Message(**prompt_message))
+
+        prompt = provider_prompt.Prompt(
+            name='default',
+            messages=prompt_messages,
+        )
+
+        if session.using_conversation is None or session.using_conversation.pipeline_uuid != pipeline_uuid:
+            conversation = provider_session.Conversation(
+                prompt=prompt,
+                messages=[],
+                pipeline_uuid=pipeline_uuid,
+                bot_uuid=bot_uuid,
+            )
+            session.conversations.append(conversation)
+            session.using_conversation = conversation
+
+        return session.using_conversation

src/langbot/pkg/provider/tools/loader.py

@@ -0,0 +1,55 @@
+from __future__ import annotations
+
+import abc
+import typing
+
+from ...core import app
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+preregistered_loaders: list[typing.Type[ToolLoader]] = []
+
+
+def loader_class(name: str):
+    """注册一个工具加载器"""
+
+    def decorator(cls: typing.Type[ToolLoader]) -> typing.Type[ToolLoader]:
+        cls.name = name
+        preregistered_loaders.append(cls)
+        return cls
+
+    return decorator
+
+
+class ToolLoader(abc.ABC):
+    """工具加载器"""
+
+    name: str = None
+
+    ap: app.Application
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+
+    async def initialize(self):
+        pass
+
+    @abc.abstractmethod
+    async def get_tools(self, bound_plugins: list[str] | None = None) -> list[resource_tool.LLMTool]:
+        """获取所有工具"""
+        pass
+
+    @abc.abstractmethod
+    async def has_tool(self, name: str) -> bool:
+        """检查工具是否存在"""
+        pass
+
+    @abc.abstractmethod
+    async def invoke_tool(self, name: str, parameters: dict) -> typing.Any:
+        """执行工具调用"""
+        pass
+
+    @abc.abstractmethod
+    async def shutdown(self):
+        """关闭工具"""
+        pass

src/langbot/pkg/provider/tools/loaders/mcp.py

@@ -0,0 +1,398 @@
+from __future__ import annotations
+
+import enum
+import typing
+from contextlib import AsyncExitStack
+import traceback
+import sqlalchemy
+import asyncio
+
+from mcp import ClientSession, StdioServerParameters
+from mcp.client.stdio import stdio_client
+from mcp.client.sse import sse_client
+
+from .. import loader
+from ....core import app
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+from ....entity.persistence import mcp as persistence_mcp
+
+
+class MCPSessionStatus(enum.Enum):
+    CONNECTING = 'connecting'
+    CONNECTED = 'connected'
+    ERROR = 'error'
+
+
+class RuntimeMCPSession:
+    """运行时 MCP 会话"""
+
+    ap: app.Application
+
+    server_name: str
+
+    server_uuid: str
+
+    server_config: dict
+
+    session: ClientSession
+
+    exit_stack: AsyncExitStack
+
+    functions: list[resource_tool.LLMTool] = []
+
+    enable: bool
+
+    # connected: bool
+    status: MCPSessionStatus
+
+    _lifecycle_task: asyncio.Task | None
+
+    _shutdown_event: asyncio.Event
+
+    _ready_event: asyncio.Event
+
+    def __init__(self, server_name: str, server_config: dict, enable: bool, ap: app.Application):
+        self.server_name = server_name
+        self.server_uuid = server_config.get('uuid', '')
+        self.server_config = server_config
+        self.ap = ap
+        self.enable = enable
+        self.session = None
+
+        self.exit_stack = AsyncExitStack()
+        self.functions = []
+
+        self.status = MCPSessionStatus.CONNECTING
+
+        self._lifecycle_task = None
+        self._shutdown_event = asyncio.Event()
+        self._ready_event = asyncio.Event()
+
+    async def _init_stdio_python_server(self):
+        server_params = StdioServerParameters(
+            command=self.server_config['command'],
+            args=self.server_config['args'],
+            env=self.server_config['env'],
+        )
+
+        stdio_transport = await self.exit_stack.enter_async_context(stdio_client(server_params))
+
+        stdio, write = stdio_transport
+
+        self.session = await self.exit_stack.enter_async_context(ClientSession(stdio, write))
+
+        await self.session.initialize()
+
+    async def _init_sse_server(self):
+        sse_transport = await self.exit_stack.enter_async_context(
+            sse_client(
+                self.server_config['url'],
+                headers=self.server_config.get('headers', {}),
+                timeout=self.server_config.get('timeout', 10),
+                sse_read_timeout=self.server_config.get('ssereadtimeout', 30),
+            )
+        )
+
+        sseio, write = sse_transport
+
+        self.session = await self.exit_stack.enter_async_context(ClientSession(sseio, write))
+
+        await self.session.initialize()
+
+    async def _lifecycle_loop(self):
+        """在后台任务中管理整个MCP会话的生命周期"""
+        try:
+            if self.server_config['mode'] == 'stdio':
+                await self._init_stdio_python_server()
+            elif self.server_config['mode'] == 'sse':
+                await self._init_sse_server()
+            else:
+                raise ValueError(f'无法识别 MCP 服务器类型: {self.server_name}: {self.server_config}')
+
+            await self.refresh()
+
+            self.status = MCPSessionStatus.CONNECTED
+
+            # 通知start()方法连接已建立
+            self._ready_event.set()
+
+            # 等待shutdown信号
+            await self._shutdown_event.wait()
+
+        except Exception as e:
+            self.status = MCPSessionStatus.ERROR
+            self.ap.logger.error(f'Error in MCP session lifecycle {self.server_name}: {e}\n{traceback.format_exc()}')
+            # 即使出错也要设置ready事件，让start()方法知道初始化已完成
+            self._ready_event.set()
+        finally:
+            # 在同一个任务中清理所有资源
+            try:
+                if self.exit_stack:
+                    await self.exit_stack.aclose()
+                self.functions.clear()
+                self.session = None
+            except Exception as e:
+                self.ap.logger.error(f'Error cleaning up MCP session {self.server_name}: {e}\n{traceback.format_exc()}')
+
+    async def start(self):
+        if not self.enable:
+            return
+
+        # 创建后台任务来管理生命周期
+        self._lifecycle_task = asyncio.create_task(self._lifecycle_loop())
+
+        # 等待连接建立或失败（带超时）
+        try:
+            await asyncio.wait_for(self._ready_event.wait(), timeout=30.0)
+        except asyncio.TimeoutError:
+            self.status = MCPSessionStatus.ERROR
+            raise Exception('Connection timeout after 30 seconds')
+
+        # 检查是否有错误
+        if self.status == MCPSessionStatus.ERROR:
+            raise Exception('Connection failed, please check URL')
+
+    async def refresh(self):
+        self.functions.clear()
+
+        tools = await self.session.list_tools()
+
+        self.ap.logger.debug(f'Refresh MCP tools: {tools}')
+
+        for tool in tools.tools:
+
+            async def func(*, _tool=tool, **kwargs):
+                result = await self.session.call_tool(_tool.name, kwargs)
+                if result.isError:
+                    raise Exception(result.content[0].text)
+                return result.content[0].text
+
+            func.__name__ = tool.name
+
+            self.functions.append(
+                resource_tool.LLMTool(
+                    name=tool.name,
+                    human_desc=tool.description,
+                    description=tool.description,
+                    parameters=tool.inputSchema,
+                    func=func,
+                )
+            )
+
+    def get_tools(self) -> list[resource_tool.LLMTool]:
+        return self.functions
+
+    def get_runtime_info_dict(self) -> dict:
+        return {
+            'status': self.status.value,
+            'tool_count': len(self.get_tools()),
+            'tools': [
+                {
+                    'name': tool.name,
+                    'description': tool.description,
+                }
+                for tool in self.get_tools()
+            ],
+        }
+
+    async def shutdown(self):
+        """关闭会话并清理资源"""
+        try:
+            # 设置shutdown事件，通知lifecycle任务退出
+            self._shutdown_event.set()
+
+            # 等待lifecycle任务完成（带超时）
+            if self._lifecycle_task and not self._lifecycle_task.done():
+                try:
+                    await asyncio.wait_for(self._lifecycle_task, timeout=5.0)
+                except asyncio.TimeoutError:
+                    self.ap.logger.warning(f'MCP session {self.server_name} shutdown timeout, cancelling task')
+                    self._lifecycle_task.cancel()
+                    try:
+                        await self._lifecycle_task
+                    except asyncio.CancelledError:
+                        pass
+
+            self.ap.logger.info(f'MCP session {self.server_name} shutdown complete')
+        except Exception as e:
+            self.ap.logger.error(f'Error shutting down MCP session {self.server_name}: {e}\n{traceback.format_exc()}')
+
+
+# @loader.loader_class('mcp')
+class MCPLoader(loader.ToolLoader):
+    """MCP 工具加载器。
+
+    在此加载器中管理所有与 MCP Server 的连接。
+    """
+
+    sessions: dict[str, RuntimeMCPSession]
+
+    _last_listed_functions: list[resource_tool.LLMTool]
+
+    _hosted_mcp_tasks: list[asyncio.Task]
+
+    def __init__(self, ap: app.Application):
+        super().__init__(ap)
+        self.sessions = {}
+        self._last_listed_functions = []
+        self._hosted_mcp_tasks = []
+
+    async def initialize(self):
+        await self.load_mcp_servers_from_db()
+
+    async def load_mcp_servers_from_db(self):
+        self.ap.logger.info('Loading MCP servers from db...')
+
+        self.sessions = {}
+
+        result = await self.ap.persistence_mgr.execute_async(sqlalchemy.select(persistence_mcp.MCPServer))
+        servers = result.all()
+
+        for server in servers:
+            config = self.ap.persistence_mgr.serialize_model(persistence_mcp.MCPServer, server)
+
+            task = asyncio.create_task(self.host_mcp_server(config))
+            self._hosted_mcp_tasks.append(task)
+
+    async def host_mcp_server(self, server_config: dict):
+        self.ap.logger.debug(f'Loading MCP server {server_config}')
+        try:
+            session = await self.load_mcp_server(server_config)
+            self.sessions[server_config['name']] = session
+        except Exception as e:
+            self.ap.logger.error(
+                f'Failed to load MCP server from db: {server_config["name"]}({server_config["uuid"]}): {e}\n{traceback.format_exc()}'
+            )
+            return
+
+        self.ap.logger.debug(f'Starting MCP server {server_config["name"]}({server_config["uuid"]})')
+        try:
+            await session.start()
+        except Exception as e:
+            self.ap.logger.error(
+                f'Failed to start MCP server {server_config["name"]}({server_config["uuid"]}): {e}\n{traceback.format_exc()}'
+            )
+            return
+
+        self.ap.logger.debug(f'Started MCP server {server_config["name"]}({server_config["uuid"]})')
+
+    async def load_mcp_server(self, server_config: dict) -> RuntimeMCPSession:
+        """加载 MCP 服务器到运行时
+
+        Args:
+            server_config: 服务器配置字典，必须包含:
+                - name: 服务器名称
+                - mode: 连接模式 (stdio/sse)
+                - enable: 是否启用
+                - extra_args: 额外的配置参数 (可选)
+        """
+
+        name = server_config['name']
+        uuid = server_config['uuid']
+        mode = server_config['mode']
+        enable = server_config['enable']
+        extra_args = server_config.get('extra_args', {})
+
+        mixed_config = {
+            'name': name,
+            'uuid': uuid,
+            'mode': mode,
+            'enable': enable,
+            **extra_args,
+        }
+
+        session = RuntimeMCPSession(name, mixed_config, enable, self.ap)
+
+        return session
+
+    async def get_tools(self, bound_mcp_servers: list[str] | None = None) -> list[resource_tool.LLMTool]:
+        all_functions = []
+
+        for session in self.sessions.values():
+            # If bound_mcp_servers is specified, only include tools from those servers
+            if bound_mcp_servers is not None:
+                if session.server_uuid in bound_mcp_servers:
+                    all_functions.extend(session.get_tools())
+            else:
+                # If no bound servers specified, include all tools
+                all_functions.extend(session.get_tools())
+
+        self._last_listed_functions = all_functions
+
+        return all_functions
+
+    async def has_tool(self, name: str) -> bool:
+        """检查工具是否存在"""
+        for session in self.sessions.values():
+            for function in session.get_tools():
+                if function.name == name:
+                    return True
+        return False
+
+    async def invoke_tool(self, name: str, parameters: dict) -> typing.Any:
+        """执行工具调用"""
+        for session in self.sessions.values():
+            for function in session.get_tools():
+                if function.name == name:
+                    self.ap.logger.debug(f'Invoking MCP tool: {name} with parameters: {parameters}')
+                    try:
+                        result = await function.func(**parameters)
+                        self.ap.logger.debug(f'MCP tool {name} executed successfully')
+                        return result
+                    except Exception as e:
+                        self.ap.logger.error(f'Error invoking MCP tool {name}: {e}\n{traceback.format_exc()}')
+                        raise
+
+        raise ValueError(f'Tool not found: {name}')
+
+    async def remove_mcp_server(self, server_name: str):
+        """移除 MCP 服务器"""
+        if server_name not in self.sessions:
+            self.ap.logger.warning(f'MCP server {server_name} not found in sessions, skipping removal')
+            return
+
+        session = self.sessions.pop(server_name)
+        await session.shutdown()
+        self.ap.logger.info(f'Removed MCP server: {server_name}')
+
+    def get_session(self, server_name: str) -> RuntimeMCPSession | None:
+        """获取指定名称的 MCP 会话"""
+        return self.sessions.get(server_name)
+
+    def has_session(self, server_name: str) -> bool:
+        """检查是否存在指定名称的 MCP 会话"""
+        return server_name in self.sessions
+
+    def get_all_server_names(self) -> list[str]:
+        """获取所有已加载的 MCP 服务器名称"""
+        return list(self.sessions.keys())
+
+    def get_server_tool_count(self, server_name: str) -> int:
+        """获取指定服务器的工具数量"""
+        session = self.get_session(server_name)
+        return len(session.get_tools()) if session else 0
+
+    def get_all_servers_info(self) -> dict[str, dict]:
+        """获取所有服务器的信息"""
+        info = {}
+        for server_name, session in self.sessions.items():
+            info[server_name] = {
+                'name': server_name,
+                'mode': session.server_config.get('mode'),
+                'enable': session.enable,
+                'tools_count': len(session.get_tools()),
+                'tool_names': [f.name for f in session.get_tools()],
+            }
+        return info
+
+    async def shutdown(self):
+        """关闭所有工具"""
+        self.ap.logger.info('Shutting down all MCP sessions...')
+        for server_name, session in list(self.sessions.items()):
+            try:
+                await session.shutdown()
+                self.ap.logger.debug(f'Shutdown MCP session: {server_name}')
+            except Exception as e:
+                self.ap.logger.error(f'Error shutting down MCP session {server_name}: {e}\n{traceback.format_exc()}')
+        self.sessions.clear()
+        self.ap.logger.info('All MCP sessions shutdown complete')

src/langbot/pkg/provider/tools/loaders/plugin.py

@@ -0,0 +1,56 @@
+from __future__ import annotations
+
+import typing
+import traceback
+
+from .. import loader
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+
+# @loader.loader_class('plugin-tool-loader')
+class PluginToolLoader(loader.ToolLoader):
+    """插件工具加载器。
+
+    本加载器中不存储工具信息，仅负责从插件系统中获取工具信息。
+    """
+
+    async def get_tools(self, bound_plugins: list[str] | None = None) -> list[resource_tool.LLMTool]:
+        # 从插件系统获取工具（内容函数）
+        all_functions: list[resource_tool.LLMTool] = []
+
+        for tool in await self.ap.plugin_connector.list_tools(bound_plugins):
+            tool_obj = resource_tool.LLMTool(
+                name=tool.metadata.name,
+                human_desc=tool.metadata.description.en_US,
+                description=tool.spec['llm_prompt'],
+                parameters=tool.spec['parameters'],
+                func=lambda parameters: {},
+            )
+            all_functions.append(tool_obj)
+
+        return all_functions
+
+    async def has_tool(self, name: str) -> bool:
+        """检查工具是否存在"""
+        for tool in await self.ap.plugin_connector.list_tools():
+            if tool.metadata.name == name:
+                return True
+        return False
+
+    async def _get_tool(self, name: str) -> resource_tool.LLMTool:
+        for tool in await self.ap.plugin_connector.list_tools():
+            if tool.metadata.name == name:
+                return tool
+        return None
+
+    async def invoke_tool(self, name: str, parameters: dict) -> typing.Any:
+        try:
+            return await self.ap.plugin_connector.call_tool(name, parameters)
+        except Exception as e:
+            self.ap.logger.error(f'执行函数 {name} 时发生错误: {e}')
+            traceback.print_exc()
+            return f'error occurred when executing function {name}: {e}'
+
+    async def shutdown(self):
+        """关闭工具"""
+        pass

src/langbot/pkg/provider/tools/toolmgr.py

@@ -0,0 +1,107 @@
+from __future__ import annotations
+
+import typing
+
+from ...core import app
+from langbot.pkg.utils import importutil
+from langbot.pkg.provider.tools import loaders
+from langbot.pkg.provider.tools.loaders import mcp as mcp_loader, plugin as plugin_loader
+import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
+
+importutil.import_modules_in_pkg(loaders)
+
+
+class ToolManager:
+    """LLM工具管理器"""
+
+    ap: app.Application
+
+    plugin_tool_loader: plugin_loader.PluginToolLoader
+    mcp_tool_loader: mcp_loader.MCPLoader
+
+    def __init__(self, ap: app.Application):
+        self.ap = ap
+
+    async def initialize(self):
+        self.plugin_tool_loader = plugin_loader.PluginToolLoader(self.ap)
+        await self.plugin_tool_loader.initialize()
+        self.mcp_tool_loader = mcp_loader.MCPLoader(self.ap)
+        await self.mcp_tool_loader.initialize()
+
+    async def get_all_tools(
+        self, bound_plugins: list[str] | None = None, bound_mcp_servers: list[str] | None = None
+    ) -> list[resource_tool.LLMTool]:
+        """获取所有函数"""
+        all_functions: list[resource_tool.LLMTool] = []
+
+        all_functions.extend(await self.plugin_tool_loader.get_tools(bound_plugins))
+        all_functions.extend(await self.mcp_tool_loader.get_tools(bound_mcp_servers))
+
+        return all_functions
+
+    async def generate_tools_for_openai(self, use_funcs: list[resource_tool.LLMTool]) -> list:
+        """生成函数列表"""
+        tools = []
+
+        for function in use_funcs:
+            function_schema = {
+                'type': 'function',
+                'function': {
+                    'name': function.name,
+                    'description': function.description,
+                    'parameters': function.parameters,
+                },
+            }
+            tools.append(function_schema)
+
+        return tools
+
+    async def generate_tools_for_anthropic(self, use_funcs: list[resource_tool.LLMTool]) -> list:
+        """为anthropic生成函数列表
+
+        e.g.
+
+        [
+          {
+            "name": "get_stock_price",
+            "description": "Get the current stock price for a given ticker symbol.",
+            "input_schema": {
+              "type": "object",
+              "properties": {
+                "ticker": {
+                  "type": "string",
+                  "description": "The stock ticker symbol, e.g. AAPL for Apple Inc."
+                }
+              },
+              "required": ["ticker"]
+            }
+          }
+        ]
+        """
+
+        tools = []
+
+        for function in use_funcs:
+            function_schema = {
+                'name': function.name,
+                'description': function.description,
+                'input_schema': function.parameters,
+            }
+            tools.append(function_schema)
+
+        return tools
+
+    async def execute_func_call(self, name: str, parameters: dict) -> typing.Any:
+        """执行函数调用"""
+
+        if await self.plugin_tool_loader.has_tool(name):
+            return await self.plugin_tool_loader.invoke_tool(name, parameters)
+        elif await self.mcp_tool_loader.has_tool(name):
+            return await self.mcp_tool_loader.invoke_tool(name, parameters)
+        else:
+            raise ValueError(f'未找到工具: {name}')
+
+    async def shutdown(self):
+        """关闭所有工具"""
+        await self.plugin_tool_loader.shutdown()
+        await self.mcp_tool_loader.shutdown()