langchain-dev-utils 1.3.2__tar.gz → 1.3.4__tar.gz

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/.gitignore

@@ -1,16 +1,16 @@
-# Python-generated files
-__pycache__/
-*.py[oc]
-build/
-dist/
-wheels/
-*.egg-info
-
-# Virtual environments
-.venv
-.env
-.benchmarks
-data/
-node_modules
-dist
+# Python-generated files
+__pycache__/
+*.py[oc]
+build/
+dist/
+wheels/
+*.egg-info
+
+# Virtual environments
+.venv
+.env
+.benchmarks
+data/
+node_modules
+dist
 site/

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/.python-version

@@ -1 +1 @@
-3.11
+3.11

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langchain-dev-utils
-Version: 1.3.2
+Version: 1.3.4
 Summary: A practical utility library for LangChain and LangGraph development
 Project-URL: Source Code, https://github.com/TBice123123/langchain-dev-utils
 Project-URL: repository, https://github.com/TBice123123/langchain-dev-utils

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/README_cn.md

@@ -1,86 +1,86 @@
-# 🦜️🧰 langchain-dev-utils
-
-<p align="center">
-    <em>🚀 专为 LangChain 和 LangGraph 开发者打造的高效工具库</em>
-</p>
-
-<p align="center">
-  📚 <a href="https://tbice123123.github.io/langchain-dev-utils/">English</a> • 
-  <a href="https://tbice123123.github.io/langchain-dev-utils/zh/">中文</a>
-</p>
-
-[![PyPI](https://img.shields.io/pypi/v/langchain-dev-utils.svg?color=%2334D058&label=pypi%20package)](https://pypi.org/project/langchain-dev-utils/)
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-[![Python](https://img.shields.io/badge/python-3.11|3.12|3.13|3.14-%2334D058)](https://www.python.org/downloads)
-[![Downloads](https://static.pepy.tech/badge/langchain-dev-utils/month)](https://pepy.tech/project/langchain-dev-utils)
-[![Documentation](https://img.shields.io/badge/docs-latest-blue)](https://tbice123123.github.io/langchain-dev-utils/zh/)
-
-> 当前为中文版，英文版请访问[English Version](https://github.com/TBice123123/langchain-dev-utils/blob/master/README.md)
-
-## ✨ 为什么选择 langchain-dev-utils？
-
-厌倦了在 LangChain 开发中编写重复代码？`langchain-dev-utils` 正是您需要的解决方案！这个轻量但功能强大的工具库专为提升 LangChain 和 LangGraph 开发体验而设计，帮助您：
-
-- ⚡ **提升开发效率** - 减少样板代码，让您专注于核心功能
-- 🧩 **简化复杂流程** - 轻松管理多模型、多工具和多智能体应用
-- 🔧 **增强代码质量** - 提高一致性和可读性，减少维护成本
-- 🎯 **加速原型开发** - 快速实现想法，更快迭代验证
-
-
-## 🎯 核心功能
-
-- **🔌 统一的模型管理** - 通过字符串指定模型提供商，轻松切换和组合不同模型
-- **💬 灵活的消息处理** - 支持思维链拼接、流式处理和消息格式化
-- **🛠️ 强大的工具调用** - 内置工具调用检测、参数解析和人工审核功能
-- **🤖 高效的 Agent 开发** - 简化智能体创建流程，扩充更多的常用中间件
-- **📊 灵活的状态图组合** - 支持串行和并行方式组合多个 StateGraph
-
-## ⚡ 快速开始
-
-**1. 安装 `langchain-dev-utils`**
-
-```bash
-pip install -U "langchain-dev-utils[standard]"
-```
-
-**2. 开始使用**
-
-```python
-from langchain.tools import tool
-from langchain_core.messages import HumanMessage
-from langchain_dev_utils.chat_models import register_model_provider, load_chat_model
-from langchain_dev_utils.agents import create_agent
-
-# 注册模型提供商
-register_model_provider("vllm", "openai-compatible", base_url="http://localhost:8000/v1")
-
-@tool
-def get_current_weather(location: str) -> str:
-    """获取指定地点的当前天气"""
-    return f"25度，{location}"
-
-# 使用字符串动态加载模型
-model = load_chat_model("vllm:qwen3-4b")
-response = model.invoke("你好")
-print(response)
-
-# 创建智能体
-agent = create_agent("vllm:qwen3-4b", tools=[get_current_weather])
-response = agent.invoke({"messages": [HumanMessage(content="今天纽约的天气如何？")]})
-print(response)
-```
-
-**获取更多的本库功能，请访问[完整文档](https://tbice123123.github.io/langchain-dev-utils/zh/)**
-
-
-## 🛠️ GitHub 仓库
-
-访问 [GitHub 仓库](https://github.com/TBice123123/langchain-dev-utils) 查看源代码和问题。
-
-
----
-
-<div align="center">
-  <p>由 ❤️ 和 ☕ 驱动开发</p>
-  <p>如果这个项目对您有帮助，请给我们一个 ⭐️</p>
+# 🦜️🧰 langchain-dev-utils
+
+<p align="center">
+    <em>🚀 专为 LangChain 和 LangGraph 开发者打造的高效工具库</em>
+</p>
+
+<p align="center">
+  📚 <a href="https://tbice123123.github.io/langchain-dev-utils/">English</a> • 
+  <a href="https://tbice123123.github.io/langchain-dev-utils/zh/">中文</a>
+</p>
+
+[![PyPI](https://img.shields.io/pypi/v/langchain-dev-utils.svg?color=%2334D058&label=pypi%20package)](https://pypi.org/project/langchain-dev-utils/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+[![Python](https://img.shields.io/badge/python-3.11|3.12|3.13|3.14-%2334D058)](https://www.python.org/downloads)
+[![Downloads](https://static.pepy.tech/badge/langchain-dev-utils/month)](https://pepy.tech/project/langchain-dev-utils)
+[![Documentation](https://img.shields.io/badge/docs-latest-blue)](https://tbice123123.github.io/langchain-dev-utils/zh/)
+
+> 当前为中文版，英文版请访问[English Version](https://github.com/TBice123123/langchain-dev-utils/blob/master/README.md)
+
+## ✨ 为什么选择 langchain-dev-utils？
+
+厌倦了在 LangChain 开发中编写重复代码？`langchain-dev-utils` 正是您需要的解决方案！这个轻量但功能强大的工具库专为提升 LangChain 和 LangGraph 开发体验而设计，帮助您：
+
+- ⚡ **提升开发效率** - 减少样板代码，让您专注于核心功能
+- 🧩 **简化复杂流程** - 轻松管理多模型、多工具和多智能体应用
+- 🔧 **增强代码质量** - 提高一致性和可读性，减少维护成本
+- 🎯 **加速原型开发** - 快速实现想法，更快迭代验证
+
+
+## 🎯 核心功能
+
+- **🔌 统一的模型管理** - 通过字符串指定模型提供商，轻松切换和组合不同模型
+- **💬 灵活的消息处理** - 支持思维链拼接、流式处理和消息格式化
+- **🛠️ 强大的工具调用** - 内置工具调用检测、参数解析和人工审核功能
+- **🤖 高效的 Agent 开发** - 简化智能体创建流程，扩充更多的常用中间件
+- **📊 灵活的状态图组合** - 支持串行和并行方式组合多个 StateGraph
+
+## ⚡ 快速开始
+
+**1. 安装 `langchain-dev-utils`**
+
+```bash
+pip install -U "langchain-dev-utils[standard]"
+```
+
+**2. 开始使用**
+
+```python
+from langchain.tools import tool
+from langchain_core.messages import HumanMessage
+from langchain_dev_utils.chat_models import register_model_provider, load_chat_model
+from langchain_dev_utils.agents import create_agent
+
+# 注册模型提供商
+register_model_provider("vllm", "openai-compatible", base_url="http://localhost:8000/v1")
+
+@tool
+def get_current_weather(location: str) -> str:
+    """获取指定地点的当前天气"""
+    return f"25度，{location}"
+
+# 使用字符串动态加载模型
+model = load_chat_model("vllm:qwen3-4b")
+response = model.invoke("你好")
+print(response)
+
+# 创建智能体
+agent = create_agent("vllm:qwen3-4b", tools=[get_current_weather])
+response = agent.invoke({"messages": [HumanMessage(content="今天纽约的天气如何？")]})
+print(response)
+```
+
+**获取更多的本库功能，请访问[完整文档](https://tbice123123.github.io/langchain-dev-utils/zh/)**
+
+
+## 🛠️ GitHub 仓库
+
+访问 [GitHub 仓库](https://github.com/TBice123123/langchain-dev-utils) 查看源代码和问题。
+
+
+---
+
+<div align="center">
+  <p>由 ❤️ 和 ☕ 驱动开发</p>
+  <p>如果这个项目对您有帮助，请给我们一个 ⭐️</p>
 </div>

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/pyproject.toml

@@ -1,47 +1,47 @@
-[project]
-name = "langchain-dev-utils"
-version = "1.3.2"
-description = "A practical utility library for LangChain and LangGraph development"
-readme = "README.md"
-authors = [{ name = "tiebingice", email = "tiebingice123@outlook.com" }]
-requires-python = ">=3.11"
-dependencies = ["langchain>=1.2.0", "langchain-core>=1.2.5", "langgraph>=1.0.0"]
-
-[project.urls]
-"Source Code" = "https://github.com/TBice123123/langchain-dev-utils"
-repository = "https://github.com/TBice123123/langchain-dev-utils"
-documentation = "https://tbice123123.github.io/langchain-dev-utils"
-
-
-[project.optional-dependencies]
-standard = ["json-repair>=0.53.1", "langchain-openai"]
-
-[build-system]
-requires = ["hatchling"]
-build-backend = "hatchling.build"
-
-[tool.hatch.build]
-exclude = ["/data", "/docs", "mkdocs.yml"]
-
-[tool.pytest.ini_options]
-asyncio_mode = "auto"
-testpaths = ["tests"]
-python_files = ["test_*.py"]
-python_functions = ["test_*"]
-
-[dependency-groups]
-dev = ["langchain-model-profiles>=0.0.5", "ruff>=0.14.5"]
-docs = ["mkdocs-material>=9.7.0", "mkdocs-static-i18n>=1.3.0"]
-tests = [
-    "python-dotenv>=1.1.1",
-    "langchain-tests>=1.0.0",
-    "langchain-deepseek>=1.0.0",
-    "langchain-qwq>=0.3.0",
-    "langchain-ollama>=1.0.0",
-    "langchain-community>=0.4.1",
-]
-
-
-[tool.ruff.lint]
-select = ["E", "F", "I", "PGH003", "T201"]
-ignore = ["E501"]
+[project]
+name = "langchain-dev-utils"
+version = "1.3.4"
+description = "A practical utility library for LangChain and LangGraph development"
+readme = "README.md"
+authors = [{ name = "tiebingice", email = "tiebingice123@outlook.com" }]
+requires-python = ">=3.11"
+dependencies = ["langchain>=1.2.0", "langchain-core>=1.2.5", "langgraph>=1.0.0"]
+
+[project.urls]
+"Source Code" = "https://github.com/TBice123123/langchain-dev-utils"
+repository = "https://github.com/TBice123123/langchain-dev-utils"
+documentation = "https://tbice123123.github.io/langchain-dev-utils"
+
+
+[project.optional-dependencies]
+standard = ["json-repair>=0.53.1", "langchain-openai"]
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build]
+exclude = ["/data", "/docs", "mkdocs.yml"]
+
+[tool.pytest.ini_options]
+asyncio_mode = "auto"
+testpaths = ["tests"]
+python_files = ["test_*.py"]
+python_functions = ["test_*"]
+
+[dependency-groups]
+dev = ["langchain-model-profiles>=0.0.5", "ruff>=0.14.5"]
+docs = ["mkdocs-material>=9.7.0", "mkdocs-static-i18n>=1.3.0"]
+tests = [
+    "python-dotenv>=1.1.1",
+    "langchain-tests>=1.0.0",
+    "langchain-deepseek>=1.0.0",
+    "langchain-qwq>=0.3.0",
+    "langchain-ollama>=1.0.0",
+    "langchain-community>=0.4.1",
+]
+
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "PGH003", "T201"]
+ignore = ["E501"]

langchain_dev_utils-1.3.4/src/langchain_dev_utils/__init__.py

@@ -0,0 +1 @@
+__version__ = "1.3.4"

langchain_dev_utils-1.3.4/src/langchain_dev_utils/agents/__init__.py

@@ -0,0 +1,4 @@
+from .factory import create_agent
+from .wrap import wrap_agent_as_tool, wrap_all_agents_as_tool
+
+__all__ = ["create_agent", "wrap_agent_as_tool", "wrap_all_agents_as_tool"]

{langchain_dev_utils-1.3.2 → langchain_dev_utils-1.3.4}/src/langchain_dev_utils/agents/middleware/handoffs.py

@@ -78,7 +78,19 @@ def _transform_agent_config(
         dict[str, AgentConfig]: The transformed agent config.
     """
 
+    new_config = {}
     for agent_name, _cfg in config.items():
+        new_config[agent_name] = {}
+
+        if "model" in _cfg:
+            new_config[agent_name]["model"] = _cfg["model"]
+        if "prompt" in _cfg:
+            new_config[agent_name]["prompt"] = _cfg["prompt"]
+        if "default" in _cfg:
+            new_config[agent_name]["default"] = _cfg["default"]
+        if "tools" in _cfg:
+            new_config[agent_name]["tools"] = _cfg["tools"]
+
         handoffs = _cfg.get("handoffs", [])
         if handoffs == "all":
             handoff_tools = [
@@ -102,8 +114,11 @@ def _transform_agent_config(
                 ]
             ]
 
-        _cfg["tools"] = [*_cfg.get("tools", []), *handoff_tools]
-    return config
+        new_config[agent_name]["tools"] = [
+            *new_config[agent_name].get("tools", []),
+            *handoff_tools,
+        ]
+    return new_config
 
 
 class HandoffAgentMiddleware(AgentMiddleware):

langchain_dev_utils-1.3.4/src/langchain_dev_utils/agents/wrap.py

@@ -0,0 +1,274 @@
+import asyncio
+from typing import Any, Awaitable, Callable, Optional, cast
+
+from langchain.tools import ToolRuntime
+from langchain_core.messages import AnyMessage, HumanMessage
+from langchain_core.tools import BaseTool, StructuredTool
+from langgraph.graph.state import CompiledStateGraph
+
+from langchain_dev_utils.message_convert import format_sequence
+
+
+def _process_input(request: str, runtime: ToolRuntime) -> str:
+    return request
+
+
+def _process_output(
+    request: str, response: list[AnyMessage], runtime: ToolRuntime
+) -> Any:
+    return response[-1].content
+
+
+def wrap_agent_as_tool(
+    agent: CompiledStateGraph,
+    tool_name: Optional[str] = None,
+    tool_description: Optional[str] = None,
+    pre_input_hooks: Optional[
+        tuple[
+            Callable[[str, ToolRuntime], str],
+            Callable[[str, ToolRuntime], Awaitable[str]],
+        ]
+        | Callable[[str, ToolRuntime], str]
+    ] = None,
+    post_output_hooks: Optional[
+        tuple[
+            Callable[[str, list[AnyMessage], ToolRuntime], Any],
+            Callable[[str, list[AnyMessage], ToolRuntime], Awaitable[Any]],
+        ]
+        | Callable[[str, list[AnyMessage], ToolRuntime], Any]
+    ] = None,
+) -> BaseTool:
+    """Wraps an agent as a tool
+
+    Args:
+        agent: The agent to wrap
+        tool_name: The name of the tool
+        tool_description: The description of the tool
+        pre_input_hooks: Hooks to run before the input is processed
+        post_output_hooks: Hooks to run after the output is processed
+
+    Returns:
+        BaseTool: The wrapped agent as a tool
+
+    Example:
+        >>> from langchain_dev_utils.agents import wrap_agent_as_tool, create_agent
+        >>>
+        >>> call_time_agent_tool = wrap_agent_as_tool(
+        ...     time_agent,
+        ...     tool_name="call_time_agent",
+        ...     tool_description="Used to invoke the time sub-agent to perform time-related tasks"
+        ... )
+        >>>
+        >>> agent = create_agent("vllm:qwen3-4b", tools=[call_time_agent_tool], name="agent")
+
+        >>> response = agent.invoke({"messages": [HumanMessage(content="What time is it now?")]})
+        >>> response
+    """
+    if agent.name is None:
+        raise ValueError("Agent name must not be None")
+
+    process_input = _process_input
+    process_input_async = _process_input
+    process_output = _process_output
+    process_output_async = _process_output
+
+    if pre_input_hooks:
+        if isinstance(pre_input_hooks, tuple):
+            process_input = pre_input_hooks[0]
+            process_input_async = pre_input_hooks[1]
+        else:
+            process_input = pre_input_hooks
+            process_input_async = pre_input_hooks
+
+    if post_output_hooks:
+        if isinstance(post_output_hooks, tuple):
+            process_output = post_output_hooks[0]
+            process_output_async = post_output_hooks[1]
+        else:
+            process_output = post_output_hooks
+            process_output_async = post_output_hooks
+
+    def call_agent(
+        request: str,
+        runtime: ToolRuntime,
+    ) -> str:
+        request = process_input(request, runtime) if process_input else request
+
+        messages = [HumanMessage(content=request)]
+        response = agent.invoke({"messages": messages})
+
+        response = process_output(request, response["messages"], runtime)
+        return response
+
+    async def acall_agent(
+        request: str,
+        runtime: ToolRuntime,
+    ) -> str:
+        if asyncio.iscoroutinefunction(process_input_async):
+            request = await process_input_async(request, runtime)
+        else:
+            request = cast(str, process_input_async(request, runtime))
+
+        messages = [HumanMessage(content=request)]
+        response = await agent.ainvoke({"messages": messages})
+
+        if asyncio.iscoroutinefunction(process_output_async):
+            response = await process_output_async(
+                request, response["messages"], runtime
+            )
+        else:
+            response = process_output(request, response["messages"], runtime)
+
+        return response
+
+    if tool_name is None:
+        tool_name = f"transfor_to_{agent.name}"
+        if not tool_name.endswith("_agent"):
+            tool_name += "_agent"
+
+    if tool_description is None:
+        tool_description = f"This tool transforms input to {agent.name}"
+
+    return StructuredTool.from_function(
+        func=call_agent,
+        coroutine=acall_agent,
+        name=tool_name,
+        description=tool_description,
+    )
+
+
+def wrap_all_agents_as_tool(
+    agents: list[CompiledStateGraph],
+    tool_name: Optional[str] = None,
+    tool_description: Optional[str] = None,
+    pre_input_hooks: Optional[
+        tuple[
+            Callable[[str, ToolRuntime], str],
+            Callable[[str, ToolRuntime], Awaitable[str]],
+        ]
+        | Callable[[str, ToolRuntime], str]
+    ] = None,
+    post_output_hooks: Optional[
+        tuple[
+            Callable[[str, list[AnyMessage], ToolRuntime], Any],
+            Callable[[str, list[AnyMessage], ToolRuntime], Awaitable[Any]],
+        ]
+        | Callable[[str, list[AnyMessage], ToolRuntime], Any]
+    ] = None,
+) -> BaseTool:
+    """Wraps all agents as single tool
+
+    Args:
+        agents: The agents to wrap
+        tool_name: The name of the tool, default to "task"
+        tool_description: The description of the tool
+        pre_input_hooks: Hooks to run before the input is processed
+        post_output_hooks: Hooks to run after the output is processed
+
+    Returns:
+        BaseTool: The wrapped agents as single tool
+
+    Example:
+        >>> from langchain_dev_utils.agents import wrap_all_agents_as_tool, create_agent
+        >>>
+        >>> call_agent_tool = wrap_all_agents_as_tool(
+        ...     [time_agent,weather_agent],
+        ...     tool_name="call_sub_agents",
+        ...     tool_description="Used to invoke the sub-agents to perform tasks"
+        ... )
+        >>>
+        >>> agent = create_agent("vllm:qwen3-4b", tools=[call_sub_agents_tool], name="agent")
+
+        >>> response = agent.invoke({"messages": [HumanMessage(content="What time is it now?")]})
+        >>> response
+    """
+    if len(agents) <= 1:
+        raise ValueError("At least more than one agent must be provided")
+
+    agents_map = {}
+
+    for agent in agents:
+        if agent.name is None:
+            raise ValueError("Agent name must not be provided")
+        if agent.name in agents_map:
+            raise ValueError("Agent name must be unique")
+        agents_map[agent.name] = agent
+
+    process_input = _process_input
+    process_input_async = _process_input
+    process_output = _process_output
+    process_output_async = _process_output
+
+    if pre_input_hooks:
+        if isinstance(pre_input_hooks, tuple):
+            process_input = pre_input_hooks[0]
+            process_input_async = pre_input_hooks[1]
+        else:
+            process_input = pre_input_hooks
+            process_input_async = pre_input_hooks
+
+    if post_output_hooks:
+        if isinstance(post_output_hooks, tuple):
+            process_output = post_output_hooks[0]
+            process_output_async = post_output_hooks[1]
+        else:
+            process_output = post_output_hooks
+            process_output_async = post_output_hooks
+
+    def call_agent(
+        agent_name: str,
+        description: str,
+        runtime: ToolRuntime,
+    ) -> str:
+        task_description = (
+            process_input(description, runtime) if process_input else description
+        )
+
+        if agent_name not in agents_map:
+            raise ValueError(f"Agent {agent_name} not found")
+
+        messages = [HumanMessage(content=task_description)]
+        response = agents_map[agent_name].invoke({"messages": messages})
+
+        response = process_output(task_description, response["messages"], runtime)
+        return response
+
+    async def acall_agent(
+        agent_name: str,
+        description: str,
+        runtime: ToolRuntime,
+    ) -> str:
+        if asyncio.iscoroutinefunction(process_input_async):
+            task_description = await process_input_async(description, runtime)
+        else:
+            task_description = cast(str, process_input_async(description, runtime))
+
+        if agent_name not in agents_map:
+            raise ValueError(f"Agent {agent_name} not found")
+
+        messages = [HumanMessage(content=task_description)]
+        response = await agents_map[agent_name].ainvoke({"messages": messages})
+
+        if asyncio.iscoroutinefunction(process_output_async):
+            response = await process_output_async(
+                task_description, response["messages"], runtime
+            )
+        else:
+            response = process_output(task_description, response["messages"], runtime)
+
+        return response
+
+    if tool_name is None:
+        tool_name = "task"
+
+    if tool_description is None:
+        tool_description = (
+            "Launch an ephemeral subagent for a task.\nAvailable agents:\n "
+            + format_sequence(list(agents_map.keys()), with_num=True)
+        )
+    return StructuredTool.from_function(
+        func=call_agent,
+        coroutine=acall_agent,
+        name=tool_name,
+        description=tool_description,
+    )

langchain_dev_utils-1.3.4/src/langchain_dev_utils/chat_models/adapters/__init__.py

@@ -0,0 +1,3 @@
+from .create_utils import create_openai_compatible_model
+
+__all__ = ["create_openai_compatible_model"]