[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-2267":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":36,"readmeContent":37,"aiSummary":38,"trendingCount":16,"starSnapshotCount":16,"syncStatus":39,"lastSyncTime":40,"discoverSource":41},2267,"agentscope","agentscope-ai\u002Fagentscope","agentscope-ai","Build and run agents you can see, understand and trust.","https:\u002F\u002Fdocs.agentscope.io\u002F",null,"Python",26702,2988,152,230,0,52,568,1743,299,45,"Apache License 2.0",false,"main",true,[27,28,29,30,31,32,33,34,35],"agent","chatbot","large-language-models","llm","llm-agent","mcp","multi-agent","multi-modal","react-agent","2026-06-12 02:00:39","\u003Cp align=\"center\">\n  \u003Cimg\n    src=\"https:\u002F\u002Fimg.alicdn.com\u002Fimgextra\u002Fi1\u002FO1CN01nTg6w21NqT5qFKH1u_!!6000000001621-55-tps-550-550.svg\"\n    alt=\"AgentScope Logo\"\n    width=\"200\"\n  \u002F>\n\u003C\u002Fp>\n\n\u003Cspan align=\"center\">\n\n[**中文主页**](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Fblob\u002Fmain\u002FREADME_zh.md) | [**Tutorial**](https:\u002F\u002Fdoc.agentscope.io\u002F) | [**Roadmap (Jan 2026 -)**](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Fblob\u002Fmain\u002Fdocs\u002Froadmap.md) | [**FAQ**](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ffaq.html)\n\n\u003C\u002Fspan>\n\n\u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fabs\u002F2402.14034\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fcs.MA-2402.14034-B31C1C?logo=arxiv&logoColor=B31C1C\"\n            alt=\"arxiv\"\n        \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fpypi.org\u002Fproject\u002Fagentscope\u002F\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython-3.10+-blue?logo=python\"\n            alt=\"pypi\"\n        \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fpypi.org\u002Fproject\u002Fagentscope\u002F\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdynamic\u002Fjson?url=https%3A%2F%2Fpypi.org%2Fpypi%2Fagentscope%2Fjson&query=%24.info.version&prefix=v&logo=pypi&label=version\"\n            alt=\"pypi\"\n        \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FeYMpfnkG8h\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F1194846673529213039?label=Discord&logo=discord\"\n            alt=\"discord\"\n        \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdoc.agentscope.io\u002F\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocs-English%7C%E4%B8%AD%E6%96%87-blue?logo=markdown\"\n            alt=\"docs\"\n        \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\".\u002FLICENSE\">\n        \u003Cimg\n            src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-Apache--2.0-black\"\n            alt=\"license\"\n        \u002F>\n    \u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n\u003Cimg src=\"https:\u002F\u002Ftrendshift.io\u002Fapi\u002Fbadge\u002Frepositories\u002F20310\" alt=\"agentscope-ai%2Fagentscope | Trendshift\" style=\"width: 250px; height: 55px;\" width=\"250\" height=\"55\"\u002F>\n\u003C\u002Fp>\n\n## What is AgentScope?\n\nAgentScope is a production-ready, easy-to-use agent framework with essential abstractions that work with rising model capability and built-in support for finetuning.\n\nWe design for increasingly agentic LLMs.\nOur approach leverages the models' reasoning and tool use abilities\nrather than constraining them with strict prompts and opinionated orchestrations.\n\n## Why use AgentScope?\n\n- **Simple**: start building your agents in 5 minutes with built-in ReAct agent, tools, skills, human-in-the-loop steering, memory, planning, realtime voice, evaluation and model finetuning\n- **Extensible**: large number of ecosystem integrations for tools, memory and observability; built-in support for MCP and A2A; message hub for flexible multi-agent orchestration and workflows\n- **Production-ready**: deploy and serve your agents locally, as serverless in the cloud, or on your K8s cluster with built-in OTel support\n\n\n\u003Cp align=\"center\">\n\u003Cimg src=\".\u002Fassets\u002Fimages\u002Fagentscope.png\" width=\"90%\" \u002F>\n\u003Cbr\u002F>\nThe AgentScope Ecosystem\n\u003C\u002Fp>\n\n\n## News\n\u003C!-- BEGIN NEWS -->\n- **[2026-04] `COMM`:** AgentScope 2.0 is on the way. [Roadmap](https:\u002F\u002Fgithub.com\u002Forgs\u002Fagentscope-ai\u002Fprojects\u002F2) | [Discussion](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Fdiscussions\u002F1441)\n- **[2026-02] `FEAT`:** Realtime Voice Agent support. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Frealtime_voice_agent) | [Multi-Agent Realtime Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_realtime) | [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ftask_realtime.html)\n- **[2026-01] `COMM`:** Biweekly Meetings launched to share ecosystem updates and development plans - join us! [Details & Schedule](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Fdiscussions\u002F1126)\n- **[2026-01] `FEAT`:** Database support & memory compression in memory module. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fshort_term_memory\u002Fmemory_compression) | [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ftask_memory.html)\n- **[2025-12] `INTG`:** A2A (Agent-to-Agent) protocol support. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fa2a_agent) | [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ftask_a2a.html)\n- **[2025-12] `FEAT`:** TTS (Text-to-Speech) support. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Ftts) | [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ftask_tts.html)\n- **[2025-11] `INTG`:** Anthropic Agent Skill support. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fagent_skill) | [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ftask_agent_skill.html)\n- **[2025-11] `RELS`:** Alias-Agent for diverse real-world tasks and Data-Juicer Agent for data processing open-sourced. [Alias-Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Falias) | [Data-Juicer Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Fdata_juicer_agent)\n- **[2025-11] `INTG`:** Agentic RL via Trinity-RFT library. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ftuner\u002Fmodel_tuning) | [Trinity-RFT](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002FTrinity-RFT)\n- **[2025-11] `INTG`:** ReMe for enhanced long-term memory. [Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Flong_term_memory\u002Freme)\n\u003C!-- END NEWS -->\n\n[More news →](.\u002Fdocs\u002FNEWS.md)\n\n## Community\n\nWelcome to join our community on\n\n| [Discord](https:\u002F\u002Fdiscord.gg\u002FeYMpfnkG8h)                                                                                         | DingTalk                                                                  |\n|----------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------|\n| \u003Cimg src=\"https:\u002F\u002Fgw.alicdn.com\u002Fimgextra\u002Fi1\u002FO1CN01hhD1mu1Dd3BWVUvxN_!!6000000000238-2-tps-400-400.png\" width=\"100\" height=\"100\"> | \u003Cimg src=\".\u002Fassets\u002Fimages\u002Fdingtalk_qr_code.png\" width=\"100\" height=\"100\"> |\n\n\u003C!-- START doctoc generated TOC please keep comment here to allow auto update -->\n\u003C!-- DON'T EDIT THIS SECTION, INSTEAD RE-RUN doctoc TO UPDATE -->\n## 📑 Table of Contents\n\n- [Quickstart](#quickstart)\n  - [Installation](#installation)\n    - [From PyPI](#from-pypi)\n    - [From source](#from-source)\n- [Example](#example)\n  - [Hello AgentScope!](#hello-agentscope)\n  - [Voice Agent](#voice-agent)\n  - [Realtime Voice Agent](#realtime-voice-agent)\n  - [Human-in-the-loop](#human-in-the-loop)\n  - [Flexible MCP Usage](#flexible-mcp-usage)\n  - [Agentic RL](#agentic-rl)\n  - [Multi-Agent Workflows](#multi-agent-workflows)\n- [Documentation](#documentation)\n- [More Examples & Samples](#more-examples--samples)\n  - [Functionality](#functionality)\n  - [Agent](#agent)\n  - [Game](#game)\n  - [Workflow](#workflow)\n  - [Evaluation](#evaluation)\n  - [Tuner](#tuner)\n- [Contributing](#contributing)\n- [License](#license)\n- [Publications](#publications)\n- [Contributors](#contributors)\n\n\u003C!-- END doctoc generated TOC please keep comment here to allow auto update -->\n\n## Quickstart\n\n### Installation\n\n> AgentScope requires **Python 3.10** or higher.\n\n#### From PyPI\n\n```bash\npip install agentscope\n```\n\nOr with uv:\n\n```bash\nuv pip install agentscope\n```\n\n#### From source\n\n```bash\n# Pull the source code from GitHub\ngit clone -b main https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope.git\n\n# Install the package in editable mode\ncd agentscope\n\npip install -e .\n# or with uv:\n# uv pip install -e .\n```\n\n\n## Example\n\n### Hello AgentScope!\n\nStart with a conversation between user and a ReAct agent 🤖 named \"Friday\"!\n\n```python\nfrom agentscope.agent import ReActAgent, UserAgent\nfrom agentscope.model import DashScopeChatModel\nfrom agentscope.formatter import DashScopeChatFormatter\nfrom agentscope.memory import InMemoryMemory\nfrom agentscope.tool import Toolkit, execute_python_code, execute_shell_command\nimport os, asyncio\n\n\nasync def main():\n    toolkit = Toolkit()\n    toolkit.register_tool_function(execute_python_code)\n    toolkit.register_tool_function(execute_shell_command)\n\n    agent = ReActAgent(\n        name=\"Friday\",\n        sys_prompt=\"You're a helpful assistant named Friday.\",\n        model=DashScopeChatModel(\n            model_name=\"qwen-max\",\n            api_key=os.environ[\"DASHSCOPE_API_KEY\"],\n            stream=True,\n        ),\n        memory=InMemoryMemory(),\n        formatter=DashScopeChatFormatter(),\n        toolkit=toolkit,\n    )\n\n    user = UserAgent(name=\"user\")\n\n    msg = None\n    while True:\n        msg = await agent(msg)\n        msg = await user(msg)\n        if msg.get_text_content() == \"exit\":\n            break\n\nasyncio.run(main())\n```\n\n### Voice Agent\n\nCreate a voice-enabled ReAct agent that can understand and respond with speech, even playing a multi-agent werewolf game with voice interactions.\n\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fc5f05254-aff6-4375-90df-85e8da95d5da\n\n\n### Realtime Voice Agent\n\nBuild a realtime voice agent with web interface that can interact with users via voice input and output.\n\n[Realtime chatbot](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Frealtime_voice_agent) | [Realtime Multi-Agent Example](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_realtime)\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F1b7b114b-e995-4586-9b3f-d3bb9fcd2558\n\n\n\n### Human-in-the-loop\n\nSupport realtime interruption in ReActAgent: conversation can be interrupted via cancellation in realtime and resumed\nseamlessly via robust memory preservation.\n\n\u003Cimg src=\".\u002Fassets\u002Fimages\u002Frealtime_steering_en.gif\" alt=\"Realtime Steering\" width=\"60%\"\u002F>\n\n### Flexible MCP Usage\n\nUse individual MCP tools as **local callable functions** to compose toolkits or wrap into a more complex tool.\n\n```python\nfrom agentscope.mcp import HttpStatelessClient\nfrom agentscope.tool import Toolkit\nimport os\n\nasync def fine_grained_mcp_control():\n    # Initialize the MCP client\n    client = HttpStatelessClient(\n        name=\"gaode_mcp\",\n        transport=\"streamable_http\",\n        url=f\"https:\u002F\u002Fmcp.amap.com\u002Fmcp?key={os.environ['GAODE_API_KEY']}\",\n    )\n\n    # Obtain the MCP tool as a **local callable function**, and use it anywhere\n    func = await client.get_callable_function(func_name=\"maps_geo\")\n\n    # Option 1: Call directly\n    await func(address=\"Tiananmen Square\", city=\"Beijing\")\n\n    # Option 2: Pass to agent as a tool\n    toolkit = Toolkit()\n    toolkit.register_tool_function(func)\n    # ...\n\n    # Option 3: Wrap into a more complex tool\n    # ...\n```\n\n### Agentic RL\n\nTrain your agentic application seamlessly with Reinforcement Learning integration. We also prepare multiple sample projects covering various scenarios:\n\n| Example                                                                                          | Description                                                 | Model                  | Training Result             |\n|--------------------------------------------------------------------------------------------------|-------------------------------------------------------------|------------------------|-----------------------------|\n| [Math Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Fmath_agent)     | Tune a math-solving agent with multi-step reasoning.        | Qwen3-0.6B             | Accuracy: 75% → 85%         |\n| [Frozen Lake](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Ffrozen_lake)   | Train an agent to navigate the Frozen Lake environment.     | Qwen2.5-3B-Instruct    | Success rate: 15% → 86%     |\n| [Learn to Ask](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Flearn_to_ask) | Tune agents using LLM-as-a-judge for automated feedback.    | Qwen2.5-7B-Instruct    | Accuracy: 47% → 92%         |\n| [Email Search](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Femail_search) | Improve tool-use capabilities without labeled ground truth. | Qwen3-4B-Instruct-2507 | Accuracy: 60%               |\n| [Werewolf Game](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Fwerewolves)  | Train agents for strategic multi-agent game interactions.   | Qwen2.5-7B-Instruct    | Werewolf win rate: 50% → 80% |\n| [Data Augment](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope-samples\u002Ftree\u002Fmain\u002Ftuner\u002Fdata_augment) | Generate synthetic training data to enhance tuning results. | Qwen3-0.6B             | AIME-24 accuracy: 20% → 60% |\n\n### Multi-Agent Workflows\n\nAgentScope provides ``MsgHub`` and pipelines to streamline multi-agent conversations, offering efficient message routing and seamless information sharing\n\n```python\nfrom agentscope.pipeline import MsgHub, sequential_pipeline\nfrom agentscope.message import Msg\nimport asyncio\n\nasync def multi_agent_conversation():\n    # Create agents\n    agent1 = ...\n    agent2 = ...\n    agent3 = ...\n    agent4 = ...\n\n    # Create a message hub to manage multi-agent conversation\n    async with MsgHub(\n        participants=[agent1, agent2, agent3],\n        announcement=Msg(\"Host\", \"Introduce yourselves.\", \"assistant\")\n    ) as hub:\n        # Speak in a sequential manner\n        await sequential_pipeline([agent1, agent2, agent3])\n        # Dynamic manage the participants\n        hub.add(agent4)\n        hub.delete(agent3)\n        await hub.broadcast(Msg(\"Host\", \"Goodbye!\", \"assistant\"))\n\nasyncio.run(multi_agent_conversation())\n```\n\n\n## Documentation\n\n- [Tutorial](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002F)\n- [FAQ](https:\u002F\u002Fdoc.agentscope.io\u002Ftutorial\u002Ffaq.html)\n- [API Docs](https:\u002F\u002Fdoc.agentscope.io\u002Fapi\u002Fagentscope.html)\n\n## More Examples & Samples\n\n### Functionality\n\n- [MCP](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fmcp)\n- [Anthropic Agent Skill](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fagent_skill)\n- [Plan](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fplan)\n- [Structured Output](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fstructured_output)\n- [RAG](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Frag)\n- [Long-Term Memory](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Flong_term_memory)\n- [Session with SQLite](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fsession_with_sqlite)\n- [Stream Printing Messages](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fstream_printing_messages)\n- [TTS](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Ftts)\n- [Code-first Deployment](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fdeployment\u002Fplanning_agent)\n- [Memory Compression](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ffunctionality\u002Fshort_term_memory\u002Fmemory_compression)\n\n### Agent\n\n- [ReAct Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Freact_agent)\n- [Voice Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fvoice_agent)\n- [Deep Research Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fdeep_research_agent)\n- [Browser-use Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fbrowser_agent)\n- [Meta Planner Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fmeta_planner_agent)\n- [A2A Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Fa2a_agent)\n- [Realtime Voice Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fagent\u002Frealtime_voice_agent)\n\n### Game\n\n- [Nine-player Werewolves](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fgame\u002Fwerewolves)\n\n### Workflow\n\n- [Multi-agent Debate](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_debate)\n- [Multi-agent Conversation](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_conversation)\n- [Multi-agent Concurrent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_concurrent)\n- [Multi-agent Realtime Conversation](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fworkflows\u002Fmultiagent_realtime)\n\n### Evaluation\n\n- [ACEBench](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Fevaluation\u002Face_bench)\n\n### Tuner\n\n- [Tune ReAct Agent](https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Ftree\u002Fmain\u002Fexamples\u002Ftuner\u002Fmodel_tuning)\n\n\n## Contributing\n\nWe welcome contributions from the community! Please refer to our [CONTRIBUTING.md](.\u002FCONTRIBUTING.md) for guidelines\non how to contribute.\n\n## License\n\nAgentScope is released under Apache License 2.0.\n\n## Publications\n\nIf you find our work helpful for your research or application, please cite our papers.\n\n- [AgentScope 1.0: A Developer-Centric Framework for Building Agentic Applications](https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.16279)\n\n- [AgentScope: A Flexible yet Robust Multi-Agent Platform](https:\u002F\u002Farxiv.org\u002Fabs\u002F2402.14034)\n\n```\n@article{agentscope_v1,\n    author  = {Dawei Gao, Zitao Li, Yuexiang Xie, Weirui Kuang, Liuyi Yao, Bingchen Qian, Zhijian Ma, Yue Cui, Haohao Luo, Shen Li, Lu Yi, Yi Yu, Shiqi He, Zhiling Luo, Wenmeng Zhou, Zhicheng Zhang, Xuguang He, Ziqian Chen, Weikai Liao, Farruh Isakulovich Kushnazarov, Yaliang Li, Bolin Ding, Jingren Zhou}\n    title   = {AgentScope 1.0: A Developer-Centric Framework for Building Agentic Applications},\n    journal = {CoRR},\n    volume  = {abs\u002F2508.16279},\n    year    = {2025},\n}\n\n@article{agentscope,\n    author  = {Dawei Gao, Zitao Li, Xuchen Pan, Weirui Kuang, Zhijian Ma, Bingchen Qian, Fei Wei, Wenhao Zhang, Yuexiang Xie, Daoyuan Chen, Liuyi Yao, Hongyi Peng, Zeyu Zhang, Lin Zhu, Chen Cheng, Hongzhu Shi, Yaliang Li, Bolin Ding, Jingren Zhou}\n    title   = {AgentScope: A Flexible yet Robust Multi-Agent Platform},\n    journal = {CoRR},\n    volume  = {abs\u002F2402.14034},\n    year    = {2024},\n}\n```\n\n## Contributors\n\nAll thanks to our contributors:\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fagentscope-ai\u002Fagentscope\u002Fgraphs\u002Fcontributors\">\n  \u003Cimg src=\"https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=agentscope-ai\u002Fagentscope&max=999&columns=12&anon=1\" \u002F>\n\u003C\u002Fa>\n","AgentScope 是一个用于构建和运行可观察、可理解和可信的代理（agents）的框架。该项目采用Python语言编写，支持大型语言模型的能力，并内置了微调支持，旨在通过利用模型的推理和工具使用能力来设计更加自主的LLM代理，而不是通过严格的提示和编排限制它们。其核心功能包括简易快速启动、高度可扩展性和生产就绪性，提供了从内置ReAct代理到多代理协调等全面的支持。适用于需要开发智能聊天机器人、多模态应用以及复杂多代理系统等场景，特别适合那些希望在不牺牲灵活性的情况下快速部署基于LLM的应用程序的开发者或团队。",2,"2026-06-11 02:49:11","top_language"]