feat(langchain): support multi-provider model dispatch

- Fix ModelService.model_info() to propagate provider field to BaseInfo,
  which also fixes the litellm direct call path defaulting to "openai"
- Modify LangChainModelAdapter.wrap_model() to dispatch to provider-specific
  LangChain chat model classes based on BaseInfo.provider:
  - anthropic -> ChatAnthropic
  - gemini/vertex_ai -> ChatGoogleGenerativeAI
  - others (openai-compatible) -> ChatOpenAI
- Add optional dependency groups: langchain-anthropic, langchain-google

类型检查通过，检查文件数 303

Change-Id: I9362b0b1e2c110007ca3aab50700c28d5d7fa7ed
Co-developed-by: Cursor <noreply@cursor.com>
Signed-off-by: congxiao.wxx <congxiao.wxx@alibaba-inc.com>

Author: congxiao-wxx

agentrun/integration/langchain/model_adapter.py

@@ -2,36 +2,106 @@
 
 将 CommonModel 包装为 LangChain BaseChatModel。"""
 
-import inspect
-import json
-from typing import Any, List, Optional
+from typing import Any, Set
 
 from agentrun.integration.langchain.message_adapter import (
     LangChainMessageAdapter,
 )
 from agentrun.integration.utils.adapter import ModelAdapter
 
+OPENAI_COMPATIBLE_PROVIDERS: Set[str] = {
+    "openai",
+    "tongyi",
+    "deepseek",
+    "moonshot",
+    "baichuan",
+    "hunyuan",
+    "minimax",
+    "spark",
+    "stepfun",
+    "wenxin",
+    "yi",
+    "zhipuai",
+    "custom",
+}
+
 
 class LangChainModelAdapter(ModelAdapter):
     """LangChain 模型适配器 / LangChain Model Adapter
 
-    将 CommonModel 包装为 LangChain BaseChatModel。"""
+    将 CommonModel 包装为 LangChain BaseChatModel。
+    根据 provider 自动选择对应的 LangChain Chat Model 类。"""
 
     def __init__(self):
         """初始化适配器，创建内部的消息适配器 / LangChain Message Adapter"""
         self._message_adapter = LangChainMessageAdapter()
 
     def wrap_model(self, common_model: Any) -> Any:
-        """包装 CommonModel 为 LangChain BaseChatModel / LangChain Model Adapter"""
+        """包装 CommonModel 为 LangChain BaseChatModel / LangChain Model Adapter
+
+        根据 BaseInfo.provider 分发到对应的 LangChain Chat Model 类：
+        - anthropic -> ChatAnthropic
+        - gemini / vertex_ai -> ChatGoogleGenerativeAI
+        - 其他（openai 兼容） -> ChatOpenAI
+        """
+        info = common_model.get_model_info()
+        provider = (info.provider or "").lower()
+
+        if provider == "anthropic":
+            return self._create_anthropic(info)
+        elif provider in ("gemini", "vertex_ai"):
+            return self._create_google(info)
+        else:
+            return self._create_openai(info)
+
+    def _create_openai(self, info: Any) -> Any:
         from langchain_openai import ChatOpenAI
 
-        info = common_model.get_model_info()  # 确保模型可用
         return ChatOpenAI(
             name=info.model,
             api_key=info.api_key,
             model=info.model,
             base_url=info.base_url,
             default_headers=info.headers,
             stream_usage=True,
-            streaming=True,  # 启用流式输出以支持 token by token
+            streaming=True,
         )
+
+    def _create_anthropic(self, info: Any) -> Any:
+        try:
+            from langchain_anthropic import ChatAnthropic
+        except ImportError as e:
+            raise ImportError(
+                "langchain-anthropic is required for Anthropic models. "
+                "Install it with: "
+                'pip install "agentrun-sdk[langchain-anthropic]"'
+            ) from e
+
+        kwargs: dict[str, Any] = {
+            "model": info.model or "",
+            "anthropic_api_key": info.api_key,
+            "default_headers": info.headers or {},
+            "streaming": True,
+        }
+        if info.base_url:
+            kwargs["anthropic_api_url"] = info.base_url
+
+        return ChatAnthropic(**kwargs)
+
+    def _create_google(self, info: Any) -> Any:
+        try:
+            from langchain_google_genai import ChatGoogleGenerativeAI
+        except ImportError as e:
+            raise ImportError(
+                "langchain-google-genai is required for Google / "
+                "Vertex AI models. Install it with: "
+                'pip install "agentrun-sdk[langchain-google]"'
+            ) from e
+
+        kwargs: dict[str, Any] = {
+            "model": info.model or "",
+            "google_api_key": info.api_key,
+            "default_headers": info.headers or {},
+        }
+
+        return ChatGoogleGenerativeAI(**kwargs)

agentrun/model/__model_service_async_template.py

@@ -231,4 +231,5 @@ def model_info(self, config: Optional[Config] = None) -> BaseInfo:
             base_url=self.provider_settings.base_url,
             model=default_model,
             headers=cfg.get_headers(),
+            provider=self.provider,
         )

agentrun/model/model_service.py

@@ -402,4 +402,5 @@ def model_info(self, config: Optional[Config] = None) -> BaseInfo:
             base_url=self.provider_settings.base_url,
             model=default_model,
             headers=cfg.get_headers(),
+            provider=self.provider,
         )

pyproject.toml

@@ -34,6 +34,16 @@ langchain = [
     "langchain-openai>=1.0.0; python_version >= '3.10'",
 ]
 
+langchain-anthropic = [
+    "langchain>=1.0.0; python_version >= '3.10'",
+    "langchain-anthropic>=0.3.0; python_version >= '3.10'",
+]
+
+langchain-google = [
+    "langchain>=1.0.0; python_version >= '3.10'",
+    "langchain-google-genai>=2.1.0; python_version >= '3.10'",
+]
+
 google-adk = [
     "google-adk>=1.18.0",
 ]