Build and run agents you can see, understand and trust.
Python agentchatbotlarge-language-modelsllmllm-agentmcpmulti-agentmulti-modalreact-agentrealtime-agentvoice-agentvoice-assistant
Support to obtain the printing messages from an agent instance as an async generator (#814)
DavdGao8845e4b2e050
diff --git a/.github/workflows/stale.yml b/.github/workflows/stale.ymlindex 3b31d25d1..26bed3d3b 100644--- a/.github/workflows/stale.yml+++ b/.github/workflows/stale.yml@@ -22,10 +22,13 @@ jobs:with:repo-token: ${{ secrets.GITHUB_TOKEN }}stale-issue-message: 'This issue is marked as stale because there has been no activity for 21 days. Remove stale label or add new comments or this issue will be closed in 3 day.'- stale-pr-message: 'This PR is marked as stale because there has been no activity for 21 days. Remove stale label or add new comments or this PR will be closed in 3 day.'close-issue-message: 'Close this stale issue.'- close-pr-message: 'Close this stale PR.'stale-issue-label: 'stale-issue'- stale-pr-label: 'stale-pr'days-before-stale: 21- days-before-close: 3\ No newline at end of file+ days-before-close: 3++ stale-pr-message: 'This PR is marked as stale because there has been no activity for 30 days. Remove stale label or add new comments or this PR will be closed in 5 day.'+ close-pr-message: 'Close this stale PR.'+ stale-pr-label: 'stale-pr'+ days-before-pr-stale: 40+ days-before-pr-close: 5\ No newline at end of filediff --git a/README.md b/README.mdindex 0f222a10b..a61a0d3b0 100644--- a/README.md+++ b/README.md@@ -351,12 +351,20 @@ as_studio- API- [API Docs](https://doc.agentscope.io/api/agentscope.html)- [Examples](https://github.com/agentscope-ai/agentscope/tree/main/examples)+ - Functional+ - [MCP](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/mcp)+ - [Plan](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/plan)+ - [Structured Output](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/structured_output)+ - [RAG](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/rag)+ - [Long-Term Memory](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/long_term_memory)+ - [Session with SQLite](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/session_with_sqlite)+ - [Stream Printing Messages](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/stream_printing_messages)- Agent- [ReAct Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/react_agent)+ - [Voice Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent/voice_agent)- [Deep Research Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent_deep_research)- [Browser-use Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent_browser)- [Meta Planner Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/meta_planner_agent)- - [Voice Agent](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent/voice_agent)- Game- [Nine-player Werewolves](https://github.com/agentscope-ai/agentscope/tree/main/examples/game/werewolves)- Workflow@@ -365,13 +373,7 @@ as_studio- [Multi-agent Concurrent](https://github.com/agentscope-ai/agentscope/tree/main/examples/workflows/multiagent_concurrent)- Evaluation- [ACEBench](https://github.com/agentscope-ai/agentscope/tree/main/examples/evaluation/ace_bench)- - Functional- - [MCP](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/mcp)- - [Plan](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/plan)- - [Structured Output](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/structured_output)- - [RAG](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/rag)- - [Long-Term Memory](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/long_term_memory)- - [Session with SQLite](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/session_with_sqlite)+## ⚖️ Licensediff --git a/README_zh.md b/README_zh.mdindex fe6f8c00f..5ac3656b8 100644--- a/README_zh.md+++ b/README_zh.md@@ -353,12 +353,20 @@ as_studio- API- [API文档](https://doc.agentscope.io/zh_CN/api/agentscope.html)- [示例](https://github.com/agentscope-ai/agentscope/tree/main/examples)+ - 功能演示+ - [MCP](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/mcp)+ - [计划](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/plan)+ - [结构化输出](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/structured_output)+ - [RAG](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/rag)+ - [长期记忆](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/long_term_memory)+ - [基于DB的会话管理](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/session_with_sqlite)+ - [流式获取打印消息](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/stream_printing_messages)- 智能体- [ReAct 智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/react_agent)+ - [语音智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent/voice_agent)- [Deep Research 智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent_deep_research)- [Browser-use 智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent_browser)- [Meta Planner 智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/meta_planner_agent)- - [语音智能体](https://github.com/agentscope-ai/agentscope/tree/main/examples/agent/voice_agent)- 游戏- [九人制狼人杀](https://github.com/agentscope-ai/agentscope/tree/main/examples/game/werewolves)- 工作流@@ -367,13 +375,7 @@ as_studio- [多智能体并发](https://github.com/agentscope-ai/agentscope/tree/main/examples/workflows/multiagent_concurrent)- 评测- [ACEBench](https://github.com/agentscope-ai/agentscope/tree/main/examples/evaluation/ace_bench)- - 功能演示- - [MCP](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/mcp)- - [计划](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/plan)- - [结构化输出](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/structured_output)- - [RAG](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/rag)- - [长期记忆](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/long_term_memory)- - [基于DB的会话管理](https://github.com/agentscope-ai/agentscope/tree/main/examples/functionality/session_with_sqlite)+## ⚖️ 许可diff --git a/docs/tutorial/en/index.rst b/docs/tutorial/en/index.rstindex 7fa644723..ff076b475 100644--- a/docs/tutorial/en/index.rst+++ b/docs/tutorial/en/index.rst@@ -67,6 +67,10 @@ Welcome to AgentScope's documentation!api/agentscope.memoryapi/agentscope.toolapi/agentscope.mcp+ api/agentscope.embedding+ api/agentscope.evaluate+ api/agentscope.plan+ api/agentscope.ragapi/agentscope.tokenapi/agentscope.tracingapi/agentscope.sessiondiff --git a/docs/tutorial/en/src/task_pipeline.py b/docs/tutorial/en/src/task_pipeline.pyindex 55f1f9971..34a41faed 100644--- a/docs/tutorial/en/src/task_pipeline.py+++ b/docs/tutorial/en/src/task_pipeline.py@@ -11,6 +11,7 @@ as syntax sugar for chaining agents together, including- **MsgHub**: a message hub for broadcasting messages among multiple agents- **sequential_pipeline** and **SequentialPipeline**: a functional and class-based implementation that chains agents in a sequential manner- **fanout_pipeline** and **FanoutPipeline**: a functional and class-based implementation that distributes the same input to multiple agents+- **stream_printing_messages**: a utility function that convert the printing messages from agent(s) into an async generator"""@@ -20,7 +21,8 @@ from agentscope.formatter import DashScopeMultiAgentFormatterfrom agentscope.message import Msgfrom agentscope.model import DashScopeChatModelfrom agentscope.agent import ReActAgent-from agentscope.pipeline import MsgHub+from agentscope.pipeline import MsgHub, stream_printing_messages+# %%# Broadcasting with MsgHub@@ -136,10 +138,11 @@ asyncio.run(check_broadcast_message())# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~# Pipeline serves as a syntax sugar for multi-agent orchestration.#-# Currently, AgentScope provides two main pipeline implementations:+# Currently, AgentScope provides three main pipeline implementations:## 1. **Sequential Pipeline**: Execute agents one by one in a predefined order# 2. **Fanout Pipeline**: Distribute the same input to multiple agents and collect their responses+# 3. **Stream Printing Messages**: Convert the printing messages from an agent into an async generator## Sequential Pipeline# ------------------------@@ -212,10 +215,45 @@ asyncio.run(check_broadcast_message())## .. tip::# By combining ``MsgHub`` and ``sequential_pipeline`` or ``fanout_pipeline``, you can create more complex workflows very easily.+#+#+# Stream Printing Messages+# -------------------------------------+# The ``stream_printing_messages`` function converts the printing messages from agent(s) into an async generator.+# It will help you to obtain the intermediate messages from the agent(s) in a streaming way.+#+# It accepts a list of agents and a coroutine task, then returns an async generator that yields tuples of ``(Msg, bool)``,+# containing the printing message during execution of the coroutine task.+#+# Note the messages with the same ``id`` are considered as the same message, and the ``last`` flag indicates whether it's the last chunk of this message.+#+# Taking the following code snippet as an example:+++async def run_example_pipeline() -> None:+ """Run an example of streaming printing messages."""+ agent = create_agent("Alice", 20, "student")++ # We disable the terminal printing to avoid messy outputs+ agent.set_console_output_enabled(False)++ async for msg, last in stream_printing_messages(+ agents=[agent],+ coroutine_task=agent(+ Msg("user", "Hello, who are you?", "user"),+ ),+ ):+ print(msg, last)+ if last:+ print()+++asyncio.run(run_example_pipeline())+# %%# Advanced Pipeline Features-# ~~~~~~~~~~~~~~~~~~~~~~~~~+# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~## Additionally, for reusability, we also provide a class-based implementation:#diff --git a/docs/tutorial/en/src/task_prompt.py b/docs/tutorial/en/src/task_prompt.pyindex 26f3d6ff1..59bc489ae 100644--- a/docs/tutorial/en/src/task_prompt.py+++ b/docs/tutorial/en/src/task_prompt.py@@ -260,7 +260,7 @@ async def run_token_counter() -> int:# We use huggingface token counter for dashscope models.token_counter = HuggingFaceTokenCounter("Qwen/Qwen2.5-VL-3B-Instruct",- use_mirror=True,+ use_mirror=False,)return await token_counter.count(formatted_message)@@ -279,7 +279,7 @@ async def run_truncated_formatter() -> None:"""Example of how to format messages with truncation."""token_counter = HuggingFaceTokenCounter(pretrained_model_name_or_path="Qwen/Qwen2.5-VL-3B-Instruct",- use_mirror=True,+ use_mirror=False,)formatter = DashScopeMultiAgentFormatter(token_counter=token_counter,diff --git a/docs/tutorial/zh_CN/index.rst b/docs/tutorial/zh_CN/index.rstindex 7fa644723..ff076b475 100644--- a/docs/tutorial/zh_CN/index.rst+++ b/docs/tutorial/zh_CN/index.rst@@ -67,6 +67,10 @@ Welcome to AgentScope's documentation!api/agentscope.memoryapi/agentscope.toolapi/agentscope.mcp+ api/agentscope.embedding+ api/agentscope.evaluate+ api/agentscope.plan+ api/agentscope.ragapi/agentscope.tokenapi/agentscope.tracingapi/agentscope.sessiondiff --git a/docs/tutorial/zh_CN/src/task_pipeline.py b/docs/tutorial/zh_CN/src/task_pipeline.pyindex 97667f828..10c752a7d 100644--- a/docs/tutorial/zh_CN/src/task_pipeline.py+++ b/docs/tutorial/zh_CN/src/task_pipeline.py@@ -11,6 +11,7 @@- **MsgHub**: 用于多个智能体之间消息的广播- **sequential_pipeline** 和 **SequentialPipeline**: 以顺序方式执行多个智能体的函数式和类式实现- **fanout_pipeline** 和 **FanoutPipeline**: 将相同输入分发给多个智能体的函数式和类式实现+- **stream_printing_messages**: 将智能体在回复过程中,调用 ``self.print`` 打印的消息转换为一个异步生成器"""@@ -20,7 +21,8 @@ from agentscope.formatter import DashScopeMultiAgentFormatterfrom agentscope.message import Msgfrom agentscope.model import DashScopeChatModelfrom agentscope.agent import ReActAgent-from agentscope.pipeline import MsgHub+from agentscope.pipeline import MsgHub, stream_printing_messages+# %%# 使用 MsgHub 进行广播@@ -134,10 +136,11 @@ asyncio.run(check_broadcast_message())# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~# 管道是 AgentScope 中多智能体编排的一种语法糖。#-# 目前,AgentScope 提供了两种主要的管道实现:+# 目前,AgentScope 提供三种管道,用于减轻开发者的负担:## 1. **顺序管道 (Sequential Pipeline)**: 按预定义顺序逐个执行智能体# 2. **扇出管道 (Fanout Pipeline)**: 将相同输入分发给多个智能体并收集它们的响应+# 3. **流式获取打印消息 (stream printing messages)**: 将智能体在回复过程中,调用 ``self.print`` 打印的消息转换为一个异步生成器## 顺序管道# ------------------------@@ -210,6 +213,39 @@ asyncio.run(check_broadcast_message())# .. tip::# 通过结合 ``MsgHub`` 和 ``sequential_pipeline`` 或 ``fanout_pipeline``,你可以非常容易地创建更复杂的工作流。+# %%+# 流式获取打印消息+# ------------------------+# ``stream_printing_messages`` 函数将智能体在回复过程中调用 ``self.print`` 打印的消息转换为一个异步生成器。+# 可以帮助开发者快速以流式方式获取智能体的中间消息。+#+# 该函数接受一个或多个智能体和一个协程任务作为输入,并返回一个异步生成器。+# 该异步生成器返回一个二元组,包含执行协程任务过程中通过 ``await self.print(...)`` 打印的消息,以及一个布尔值,表示该消息是否为一组流式消息中的最后一个。+#+# 需要注意的是,生成器返回的元组中,布尔值表示该消息是否为一组流式消息中的最后一个,而非此次智能体调用的最后一条消息。+++async def run_example_pipeline() -> None:+ """运行流式打印消息的示例。"""+ agent = create_agent("Alice", 20, "student")++ # 我们关闭agent的终端打印以避免输出混乱+ agent.set_console_output_enabled(False)++ async for msg, last in stream_printing_messages(+ agents=[agent],+ coroutine_task=agent(+ Msg("user", "你好,你是谁?", "user"),+ ),+ ):+ print(msg, last)+ if last:+ print()+++asyncio.run(run_example_pipeline())++# %%# 高级管道特性# ~~~~~~~~~~~~~~~~~~~~~~~~~diff --git a/docs/tutorial/zh_CN/src/task_prompt.py b/docs/tutorial/zh_CN/src/task_prompt.pyindex fca0403a7..1e4196cef 100644--- a/docs/tutorial/zh_CN/src/task_prompt.py+++ b/docs/tutorial/zh_CN/src/task_prompt.py@@ -260,7 +260,7 @@ async def run_token_counter() -> int:# 我们使用 huggingface token 计数器用于 dashscope 模型。token_counter = HuggingFaceTokenCounter("Qwen/Qwen2.5-VL-3B-Instruct",- use_mirror=True,+ use_mirror=False,)return await token_counter.count(formatted_message)@@ -279,7 +279,7 @@ async def run_truncated_formatter() -> None:"""带截断的消息格式化示例。"""token_counter = HuggingFaceTokenCounter(pretrained_model_name_or_path="Qwen/Qwen2.5-VL-3B-Instruct",- use_mirror=True,+ use_mirror=False,)formatter = DashScopeMultiAgentFormatter(token_counter=token_counter,diff --git a/examples/functionality/stream_printing_messages/README.md b/examples/functionality/stream_printing_messages/README.mdnew file mode 100644index 000000000..be04293cb--- /dev/null+++ b/examples/functionality/stream_printing_messages/README.md@@ -0,0 +1,28 @@+# Stream Printing Messages++The AgentScope agent is designed to communicate with the user and the other agents by passing messages explicitly.+However, we notice the requirements that obtain the printing messages from the agent in a streaming manner.+Therefore, in example we demonstrate how to gather and yield the printing messages from a single agent and+multi-agent systems in a streaming manner.+++## Quick Start++Run the following command to see the streaming printing messages from the agent.+Note the messages with the same ID are the chunks of the same message in accumulated manner.++- For single-agent:++```bash+python single_agent.py+```++- For multi-agent:++```bash+python multi_agent.py+```++> Note: The example is built with DashScope chat model. If you want to change the model in this example, don't forget+> to change the formatter at the same time! The corresponding relationship between built-in models and formatters are+> list in [our tutorial](https://doc.agentscope.io/tutorial/task_prompt.html#id1)\ No newline at end of filediff --git a/examples/functionality/stream_printing_messages/multi_agent.py b/examples/functionality/stream_printing_messages/multi_agent.pynew file mode 100644index 000000000..4925af6e9--- /dev/null+++ b/examples/functionality/stream_printing_messages/multi_agent.py@@ -0,0 +1,62 @@+# -*- coding: utf-8 -*-+"""Example for gather the printing messages from multiple agents."""+import asyncio+import os++from agentscope.agent import ReActAgent+from agentscope.formatter import DashScopeMultiAgentFormatter+from agentscope.message import Msg+from agentscope.model import DashScopeChatModel+from agentscope.pipeline import MsgHub, stream_printing_messages+++def create_agent(name: str) -> ReActAgent:+ """Create an agent with the given name."""+ return ReActAgent(+ name=name,+ sys_prompt=f"You are a student named {name}.",+ model=DashScopeChatModel(+ api_key=os.environ["DASHSCOPE_API_KEY"],+ model_name="qwen-max",+ stream=False, # close streaming for simplicity+ ),+ formatter=DashScopeMultiAgentFormatter(),+ )+++async def workflow(+ alice: ReActAgent,+ bob: ReActAgent,+ charlie: ReActAgent,+) -> None:+ """The example workflow for multiple agents."""+ async with MsgHub(+ participants=[alice, bob, charlie],+ announcement=Msg(+ "user",+ "Alice, Bob and Charlie, welcome to the meeting! Let's "+ "meet each other first.",+ "user",+ ),+ ):+ # agent speaks in turn+ await alice()+ await bob()+ await charlie()+++async def main() -> None:+ """The main entry for the example."""+ # Create agents+ alice, bob, charlie = [+ create_agent(_) for _ in ["Alice", "Bob", "Charlie"]+ ]++ async for msg, last in stream_printing_messages(+ agents=[alice, bob, charlie],+ coroutine_task=workflow(alice, bob, charlie),+ ):+ print(msg, last)+++asyncio.run(main())diff --git a/examples/functionality/stream_printing_messages/single_agent.py b/examples/functionality/stream_printing_messages/single_agent.pynew file mode 100644index 000000000..7fa6afe1a--- /dev/null+++ b/examples/functionality/stream_printing_messages/single_agent.py@@ -0,0 +1,62 @@+# -*- coding: utf-8 -*-+"""The example demonstrating how to obtain the messages from the agent in a+streaming way."""+import asyncio+import os++from agentscope.agent import ReActAgent+from agentscope.formatter import DashScopeChatFormatter+from agentscope.memory import InMemoryMemory+from agentscope.message import Msg+from agentscope.model import DashScopeChatModel+from agentscope.pipeline import stream_printing_messages+from agentscope.tool import (+ Toolkit,+ execute_shell_command,+ view_text_file,+ execute_python_code,+)+++async def main() -> None:+ """The main function."""+ toolkit = Toolkit()+ toolkit.register_tool_function(execute_shell_command)+ toolkit.register_tool_function(execute_python_code)+ toolkit.register_tool_function(view_text_file)++ agent = ReActAgent(+ name="Friday",+ sys_prompt="You are a helpful assistant named Friday.",+ # Change the model and formatter together if you want to try other+ # models+ model=DashScopeChatModel(+ api_key=os.environ.get("DASHSCOPE_API_KEY"),+ model_name="qwen-max",+ enable_thinking=False,+ stream=True,+ ),+ formatter=DashScopeChatFormatter(),+ toolkit=toolkit,+ memory=InMemoryMemory(),+ )++ # Prepare a user message+ user_msg = Msg(+ "user",+ "Hi! Who are you?",+ "user",+ )++ # We disable the terminal printing to avoid messy outputs+ agent.set_console_output_enabled(False)++ # obtain the printing messages from the agent in a streaming way+ async for msg, last in stream_printing_messages(+ agents=[agent],+ coroutine_task=agent(user_msg),+ ):+ print(msg, last)+++asyncio.run(main())diff --git a/src/agentscope/_version.py b/src/agentscope/_version.pyindex a92a89fb3..757782df4 100644--- a/src/agentscope/_version.py+++ b/src/agentscope/_version.py@@ -1,4 +1,4 @@# -*- coding: utf-8 -*-"""The version of agentscope."""-__version__ = "1.0.4"+__version__ = "1.0.5dev"diff --git a/src/agentscope/agent/_agent_base.py b/src/agentscope/agent/_agent_base.pyindex dac803bd2..c90c07516 100644--- a/src/agentscope/agent/_agent_base.py+++ b/src/agentscope/agent/_agent_base.py@@ -2,12 +2,13 @@"""The agent base class in agentscope."""import asyncioimport json-from asyncio import Task+from asyncio import Task, Queuefrom collections import OrderedDictfrom typing import Callable, Anyimport base64import shortuuidimport numpy as np+from typing_extensions import deprecatedfrom ._agent_meta import _AgentMetafrom .._logging import logger@@ -167,6 +168,11 @@ class AgentBase(StateModule, metaclass=_AgentMeta):# output of the agent, e.g., in a production environment.self._disable_console_output: bool = False+ # The streaming message queue used to export the messages as a+ # generator+ self._disable_msg_queue: bool = True+ self.msg_queue = None+async def observe(self, msg: Msg | list[Msg] | None) -> None:"""Receive the given message(s) without generating a reply.@@ -197,6 +203,9 @@ class AgentBase(StateModule, metaclass=_AgentMeta):Whether this is the last one in streaming messages. Fornon-streaming message, this should always be `True`."""+ if not self._disable_msg_queue:+ await self.msg_queue.put((msg, last))+if self._disable_console_output:return@@ -622,7 +631,46 @@ class AgentBase(StateModule, metaclass=_AgentMeta):else:self._subscribers.pop(msghub_name)+ @deprecated("Please use set_console_output_enabled() instead.")def disable_console_output(self) -> None:"""This function will disable the console output of the agent, e.g.in a production environment to avoid messy logs."""self._disable_console_output = True++ def set_console_output_enabled(self, enabled: bool) -> None:+ """Enable or disable the console output of the agent. E.g. in a+ production environment, you may want to disable the console output to+ avoid messy logs.++ Args:+ enabled (`bool`):+ If `True`, enable the console output. If `False`, disable+ the console output.+ """+ self._disable_console_output = not enabled++ def set_msg_queue_enabled(+ self,+ enabled: bool,+ queue: Queue | None = None,+ ) -> None:+ """Enable or disable the message queue for streaming outputs.++ Args:+ enabled (`bool`):+ If `True`, enable the message queue to allow streaming+ outputs. If `False`, disable the message queue.+ queue (`Queue | None`, optional):+ The queue instance that will be used to initialize the+ message queue when `enable` is `True`.+ """+ if enabled:+ if queue is None:+ if self.msg_queue is None:+ self.msg_queue = asyncio.Queue(maxsize=100)+ else:+ self.msg_queue = queue+ else:+ self.msg_queue = None++ self._disable_msg_queue = not enableddiff --git a/src/agentscope/pipeline/__init__.py b/src/agentscope/pipeline/__init__.pyindex 9e0f5b801..9ab9b581e 100644--- a/src/agentscope/pipeline/__init__.py+++ b/src/agentscope/pipeline/__init__.py@@ -4,7 +4,11 @@ complex workflows and multi-agent conversations."""from ._msghub import MsgHubfrom ._class import SequentialPipeline, FanoutPipeline-from ._functional import sequential_pipeline, fanout_pipeline+from ._functional import (+ sequential_pipeline,+ fanout_pipeline,+ stream_printing_messages,+)__all__ = ["MsgHub",@@ -12,4 +16,5 @@ __all__ = ["sequential_pipeline","FanoutPipeline","fanout_pipeline",+ "stream_printing_messages",]diff --git a/src/agentscope/pipeline/_functional.py b/src/agentscope/pipeline/_functional.pyindex a6fd374cb..e9e96f720 100644--- a/src/agentscope/pipeline/_functional.py+++ b/src/agentscope/pipeline/_functional.py@@ -2,7 +2,7 @@"""Functional counterpart for Pipeline"""import asynciofrom copy import deepcopy-from typing import Any+from typing import Any, AsyncGenerator, Tuple, Coroutinefrom ..agent import AgentBasefrom ..message import Msg@@ -102,3 +102,68 @@ async def fanout_pipeline(return await asyncio.gather(*tasks)else:return [await agent(deepcopy(msg), **kwargs) for agent in agents]+++async def stream_printing_messages(+ agents: list[AgentBase],+ coroutine_task: Coroutine,+ end_signal: str = "[END]",+) -> AsyncGenerator[Tuple[Msg, bool], None]:+ """This pipeline will gather the printing messages from agents when+ execute the given coroutine task, and yield them one by one.+ Only the messages that are printed by `await self.print(msg)` in the agent+ will be forwarded to the message queue and yielded by this pipeline.++ .. note:: The boolean in the yielded tuple indicates whether the message+ is the last **chunk** for a streaming message, not the last message+ returned by the agent. That means, there'll be multiple tuples with+ `is_last_chunk=True` if the agent prints multiple messages.++ .. note:: The messages with the same ``id`` is considered as the same+ message, e.g., the chunks of a streaming message.++ Args:+ agents (`list[AgentBase]`):+ A list of agents whose printing messages will be gathered and+ yielded.+ coroutine_task (`Coroutine`):+ The coroutine task to be executed. This task should involve the+ execution of the provided agents, so that their printing messages+ can be captured and yielded.+ end_signal (`str`, defaults to `"[END]"`):+ A special signal to indicate the end of message streaming. When+ this signal is received from the message queue, the generator will+ stop yielding messages and exit the loop.++ Returns:+ `AsyncGenerator[Tuple[Msg, bool], None]`:+ An async generator that yields tuples of (message, is_last_chunk).+ The `is_last_chunk` boolean indicates whether the message is the+ last chunk in a streaming message.+ """++ # Enable the message queue to get the intermediate messages+ queue = asyncio.Queue()+ for agent in agents:+ # Use one queue to gather messages from all agents+ agent.set_msg_queue_enabled(True, queue)++ # Execute the agent asynchronously+ task = asyncio.create_task(coroutine_task)++ if task.done():+ await queue.put(end_signal)+ else:+ task.add_done_callback(lambda _: queue.put_nowait(end_signal))++ # Receive the messages from the agent's message queue+ while True:+ # The message obj, and a boolean indicating whether it's the last chunk+ # in a streaming message+ printing_msg = await queue.get()++ # End the loop when the message is None+ if isinstance(printing_msg, str) and printing_msg == end_signal:+ break++ yield printing_msgdiff --git a/tests/pipeline_test.py b/tests/pipeline_test.pyindex ce2aca09a..af6429b1b 100644--- a/tests/pipeline_test.py+++ b/tests/pipeline_test.py@@ -9,6 +9,7 @@ from agentscope.pipeline import (FanoutPipeline,sequential_pipeline,fanout_pipeline,+ stream_printing_messages,)from agentscope.agent import AgentBase@@ -41,6 +42,50 @@ class AddAgent(AgentBase):"""Handle interrupt"""+class StreamAgent(AgentBase):+ """Add agent class."""++ def __init__(self) -> None:+ """Initialize the agent"""+ super().__init__()+ self.name = "Stream"++ async def reply(self) -> Msg | None:+ """Reply function"""+ await self.print(+ Msg(+ self.name,+ "123",+ "user",+ ),+ )+ await self.print(+ Msg(+ "user",+ "456",+ "user",+ ),+ )+ await self.print(+ Msg(+ self.name,+ "789",+ "user",+ ),+ )+ return None++ async def observe(self, msg: Msg | list[Msg] | None) -> None:+ """Observe function"""++ async def handle_interrupt(+ self,+ *args: Any,+ **kwargs: Any,+ ) -> Msg:+ """Handle interrupt"""++class MultAgent(AgentBase):"""Mult agent class."""@@ -299,3 +344,35 @@ class PipelineTest(IsolatedAsyncioTestCase):self.assertEqual(len(res), 2)self.assertIsNone(res[0])self.assertIsNone(res[1])++ async def test_stream_printing_messages(self) -> None:+ """Test stream_printing_messages function"""++ agent = StreamAgent()++ i = 0+ async for msg, last in stream_printing_messages(+ [agent],+ agent(),+ ):+ self.assertTrue(last)++ if i == 0:+ self.assertEqual(+ msg.content,+ "123",+ )++ elif i == 1:+ self.assertEqual(+ msg.content,+ "456",+ )++ elif i == 2:+ self.assertEqual(+ msg.content,+ "789",+ )++ i += 1