{"id":76458,"url":"https://github.com/yzfly/awesome-agi-agents","name":"awesome-agi-agents","description":"🤖 Awesome list of AGI Agents. Agents 精选资源合集.","projects_count":69,"last_synced_at":"2026-10-02T20:00:55.159Z","repository":{"id":187500960,"uuid":"676854480","full_name":"yzfly/Awesome-AGI-Agents","owner":"yzfly","description":"🤖 Awesome list of AGI Agents. Agents 精选资源合集.","archived":false,"fork":false,"pushed_at":"2026-08-25T13:49:26.000Z","size":30,"stargazers_count":559,"open_issues_count":2,"forks_count":47,"subscribers_count":7,"default_branch":"main","last_synced_at":"2026-09-12T23:19:03.253Z","etag":null,"topics":["agents","agents-art","agi","agi-agents","ai","ai-agents","artificial-intelligence","chatgpt","claude-ai","deep-learning","gpt-4","gpt-5","llm"],"latest_commit_sha":null,"homepage":"https://agents.art","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/yzfly.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"claude":null,"gemini":null,"cursor":null,"copilot":null,"dco":null,"cla":null,"disclosure":null}},"created_at":"2023-08-10T06:58:35.000Z","updated_at":"2026-09-07T06:33:54.000Z","dependencies_parsed_at":"2026-09-12T23:00:44.854Z","dependency_job_id":"e344878c-f1ce-4903-9e8c-a8fc92679556","html_url":"https://github.com/yzfly/Awesome-AGI-Agents","commit_stats":null,"previous_names":["yzfly/awesome-agents","yzfly/awesome-agi-agents"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/yzfly/Awesome-AGI-Agents","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yzfly%2FAwesome-AGI-Agents","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yzfly%2FAwesome-AGI-Agents/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yzfly%2FAwesome-AGI-Agents/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yzfly%2FAwesome-AGI-Agents/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/yzfly","download_url":"https://codeload.github.com/yzfly/Awesome-AGI-Agents/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yzfly%2FAwesome-AGI-Agents/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":342542704,"owners_count":37944735,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-08-22T15:14:58.755Z","status":"online","status_checked_at":"2026-10-02T02:00:07.254Z","response_time":104,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"created_at":"2024-11-08T06:08:15.240Z","updated_at":"2026-10-02T20:00:55.160Z","primary_language":null,"list_of_lists":false,"displayable":true,"categories":["论文","前沿项目","Curated List","Agents 开发平台","2024-2026 重要项目","文章和视频","Star History"],"sub_categories":[],"readme":"# Awesome-AGI-Agents\n\nAgents (智能体) 精选资源合集，持续更新中\n\n## 目录\n\n- [文章和视频](#文章和视频)\n- [论文](#论文)\n- [2024-2026 重要项目](#2024-2026-重要项目)\n- [前沿项目](#前沿项目)\n- [Agents 开发平台](#agents-开发平台)\n- [Curated List](#curated-list)\n- [Star History](#star-history)\n\n## 文章和视频\n\n|名称|简介|备注|\n|---|---|---|\n|[AI Agents大爆发：软件2.0雏形初现，OpenAI的下一步](https://mp.weixin.qq.com/s/Jb8HBbaKYXXxTSQOBsP5Wg)|Lilian Weng 的个人博客文章，Lilian 现在是 OpenAI 的 Head of Safety Systems，之前还领导过 OpenAI 的 Applied AI 团队。AI Agent 被认为是 OpenAI 发力的下一个方向。OpenAI 的联合创始人 Andrej Karpathy 在近期的一次公开活动上提到“相比模型训练方法，OpenAI 内部目前更关注 Agent 领域的变化，每当有新的 AI Agents 论文出来的时候，内部都会很兴奋并且认真地讨论”，而在更早之前，Andrej  还评价 AutoGPT 是 Prompt Engineering 下一阶段的探索方向。|[英文原文](https://lilianweng.github.io/posts/2023-06-23-agent/)|\n|[《综述：全新大语言模型驱动的Agent》——4.5万字详细解读复旦NLP和米哈游最新Agent Survey](https://zhuanlan.zhihu.com/p/656676717)|复旦NLP团队和米哈游一起出的《TITLE：The Rise and Potential of Large Language Model Based Agents: A Survey》论文一个翻译版本，作者同时对部分内容进行了删繁就简，总结概括。|[论文原文](https://arxiv.org/pdf/2309.07864v1.pdf)|\n|[AI Agent的千亿美金问题：如何重构10亿知识工作职业，掀起软件生产革命？](https://mp.weixin.qq.com/s/JYu_oXWbWbasT1fcBRo-cA)|作者认为 agent 产品需要具备的特性是，要给产品设计者和用户提供干预空间。目前实践中最具代表性的有两类：中间层的 Agent Framework 和垂直领域的 Vertical Agent。|-|\n|[LangChain Agents - Joining Tools and Chains with Decisions](https://www.youtube.com/watch?v=ziu87EXZVUE\u0026ab_channel=SamWitteveen)|LangChain Agents--将工具和任务链与决策结合起来|英文 Youtube 视频，LangChain 项目官方对预置的 agents 介绍。|-|\n\n\n## 论文\n|名称|简介|备注|\n|---|---|---|\n|[LLM-Agent-Survey](https://github.com/Paitesanshi/LLM-Agent-Survey)| LLM-based Autonomous Agents 综述| ![GitHub Repo stars](https://badgen.net/github/stars/Paitesanshi/LLM-Agent-Survey)|\n|[LLM-Agent-Paper-List](https://github.com/WooooDyy/LLM-Agent-Paper-List)| \"The Rise and Potential of Large Language Model Based Agents: A Survey\"综述论文的论文列表| ![GitHub Repo stars](https://badgen.net/github/stars/WooooDyy/LLM-Agent-Paper-List)|\n|[LLMAgentPapers](https://github.com/zjunlp/LLMAgentPapers)| LLM-based Autonomous Agents 综述| ![GitHub Repo stars](https://badgen.net/github/stars/zjunlp/LLMAgentPapers)|\n|[Toolformer: Language Models Can Teach Themselves to Use Tools](https://arxiv.org/abs/2302.04761)|大语言模型可以学会使用工具|-|\n|[HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](https://arxiv.org/abs/2303.17580)|用ChatGPT作为控制器，连接HuggingFace社区中的各种AI模型，完成多模态复杂任务。整个过程，只需要做的是：用自然语言将你的需求输出。|[知乎中文讨论](https://www.zhihu.com/question/594533230/answer/2975525808)|\n|[ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs](https://arxiv.org/abs/2307.16789)|研究人员设计了一个评测 LLM 使用工具能力的 Benchmark（基准）—— LLMBench，以及一个针对该场景的数据构建、模型训练、评测的框架—— ToolLLM。|[[知乎文章](https://zhuanlan.zhihu.com/p/649277843)], [[GitHub 开源地址](https://github.com/OpenBMB/ToolBench)]|\n|[AgentBench: Evaluating LLMs as Agents](https://arxiv.org/abs/2308.03688)| AgentBench 是首个旨在评估 LLM 在各种不同环境中作为智能体的基准。它包含 8 种不同的环境，可以更全面地评估 LLM 在各种场景中作为自主智能体运行的能力。|[GitHub 开源地址](https://github.com/THUDM/AgentBench)|\n|[ClawBench: Can AI Agents Complete Everyday Online Tasks?](https://arxiv.org/abs/2604.08523)|面向真实线上任务的浏览器智能体评测框架，包含 153 个日常任务并覆盖 144 个真实网站；通过隔离容器、多种可选 agent harness，以及回放、动作、网络和消息轨迹支持可复现实验与失效分析。|[项目主页](https://claw-bench.com/) · [GitHub](https://github.com/TIGER-AI-Lab/ClawBench)|\n\n\n## 2024-2026 重要项目\n\n本仓库早期内容主要覆盖 2023 年的 Agent 生态。以下为 2024-2026 年间崛起并经验证仍活跃的重要项目与协议（2026-08 更新：新增 deepseek-harness、Codex CLI、OpenClaw、Hermes、pi、Agent Skills）。\n\n|名称|Stars|简介|\n|---|---|---|\n|[LangGraph](https://github.com/langchain-ai/langgraph)|![GitHub Repo stars](https://badgen.net/github/stars/langchain-ai/langgraph)|LangChain 团队推出的有状态、多角色 Agent 编排框架，以图 (Graph) 的方式构建可控、可循环的 Agent 工作流。|\n|[CrewAI](https://github.com/crewAIInc/crewAI)|![GitHub Repo stars](https://badgen.net/github/stars/crewAIInc/crewAI)|面向角色协作的多 Agent 框架，让多个扮演不同角色的智能体像团队一样协同完成复杂任务。|\n|[OpenAI Agents SDK](https://github.com/openai/openai-agents-python)|![GitHub Repo stars](https://badgen.net/github/stars/openai/openai-agents-python)|OpenAI 官方推出的轻量级多 Agent 编排 SDK，内置 handoffs, guardrails 与 tracing，是 Swarm 的生产级继任者。|\n|[OpenHands](https://github.com/OpenHands/OpenHands)|![GitHub Repo stars](https://badgen.net/github/stars/OpenHands/OpenHands)|原 OpenDevin，能够编写代码, 运行命令, 浏览网页的自主软件工程 Agent 平台。|\n|[Orkas](https://github.com/Orkas-AI/Orkas)|![GitHub Repo stars](https://badgen.net/github/stars/Orkas-AI/Orkas)|开源、本地优先的 AI 团队桌面应用，由 Commander 将目标转成可执行计划，并在一个对话中协调专业 Agent 并行或串行完成复杂工作。|\n|[SWE-agent](https://github.com/SWE-agent/SWE-agent)|![GitHub Repo stars](https://badgen.net/github/stars/SWE-agent/SWE-agent)|普林斯顿大学出品，让语言模型自主修复 GitHub Issue 的 Agent，在 SWE-bench 上表现出色。|\n|[Model Context Protocol (MCP)](https://github.com/modelcontextprotocol)|![GitHub Repo stars](https://badgen.net/github/stars/modelcontextprotocol/modelcontextprotocol)|Anthropic 提出的开放协议，为 LLM 与外部工具, 数据源的连接提供统一标准，已成为 Agent 生态事实标准。|\n|[Browser Use](https://github.com/browser-use/browser-use)|![GitHub Repo stars](https://badgen.net/github/stars/browser-use/browser-use)|让 AI Agent 像人一样操作浏览器，是当下最流行的网页自动化 Agent 工具之一。|\n|[agent-qa](https://github.com/vostride/agent-qa)|![GitHub Repo stars](https://badgen.net/github/stars/vostride/agent-qa)|面向软件团队的开源自进化 QA Agent，支持用自然语言编写并运行 Web 与移动端测试，通过跨运行记忆适应 UI 变化并发现回归问题。|\n|[smolagents](https://github.com/huggingface/smolagents)|![GitHub Repo stars](https://badgen.net/github/stars/huggingface/smolagents)|Hugging Face 推出的极简 Agent 库，核心仅约一千行代码，主打以代码形式编写 Agent 动作。|\n|[Pydantic AI](https://github.com/pydantic/pydantic-ai)|![GitHub Repo stars](https://badgen.net/github/stars/pydantic/pydantic-ai)|由 Pydantic 团队打造的 Agent 框架，强调类型安全与结构化输出，将 Pydantic 的工程体验带入 GenAI 开发。|\n|[A2A (Agent2Agent)](https://github.com/a2aproject/A2A)|![GitHub Repo stars](https://badgen.net/github/stars/a2aproject/A2A)|Google 发起的开放协议 (现由 Linux 基金会托管)，用于不同框架, 不同厂商的 Agent 之间互联互通。|\n|[Claude Code](https://github.com/anthropics/claude-code)|![GitHub Repo stars](https://badgen.net/github/stars/anthropics/claude-code)|Anthropic 官方推出的终端编程 Agent，可在命令行中理解代码库、编辑文件、运行命令并完成复杂工程任务，是 2025 年最具影响力的 coding agent 之一。|\n|[Gemini CLI](https://github.com/google-gemini/gemini-cli)|![GitHub Repo stars](https://badgen.net/github/stars/google-gemini/gemini-cli)|Google 官方开源的终端 AI Agent，将 Gemini 模型带入命令行，支持代码理解、工具调用与 MCP，采用 Apache-2.0 协议。|\n|[Cline](https://github.com/cline/cline)|![GitHub Repo stars](https://badgen.net/github/stars/cline/cline)|VS Code 中的自主编程 Agent，可创建/编辑文件、执行命令、使用浏览器，并原生支持 MCP，是最流行的开源 IDE 编程助手之一。|\n|[Google ADK](https://github.com/google/adk-python)|![GitHub Repo stars](https://badgen.net/github/stars/google/adk-python)|Google 开源的 Agent Development Kit，模型与部署无关的代码优先 Agent 开发框架，为 Google 自家 Agent 产品提供支撑，并原生支持 A2A/MCP。|\n|[Microsoft Agent Framework](https://github.com/microsoft/agent-framework)|![GitHub Repo stars](https://badgen.net/github/stars/microsoft/agent-framework)|微软推出的新一代开源 Agent 框架，统一并继承 AutoGen 与 Semantic Kernel，面向生产级多 Agent 编排与工作流。|\n|[AG-UI](https://github.com/ag-ui-protocol/ag-ui)|![GitHub Repo stars](https://badgen.net/github/stars/ag-ui-protocol/ag-ui)|Agent-User Interaction Protocol，用于连接后端 Agent 与前端应用的开放协议，与 MCP(接工具)、A2A(接 Agent)互补，专注 Agent 与用户界面的实时交互。|\n|[deepseek-harness (dsh)](https://github.com/deepseek-ai/deepseek-harness)|![GitHub Repo stars](https://badgen.net/github/stars/deepseek-ai/deepseek-harness)|DeepSeek 官方开源 Agent 框架（2026-08-13 发布，MIT，两周破 19 万 star）：\"一切皆插件\"，模型、工具、会话、沙箱、循环与 UI 均可替换；内置 MCP 客户端，兼容 Claude Code / Codex 的 Hook 协议与 Agent Skills。技能精选见 [awesome-dsh-skills](https://github.com/yzfly/awesome-dsh-skills)。|\n|[OpenAI Codex CLI](https://github.com/openai/codex)|![GitHub Repo stars](https://badgen.net/github/stars/openai/codex)|OpenAI 官方开源终端编码 Agent，基于 Responses API，支持 MCP 与 Agent Skills，可接入 DeepSeek-V4 等第三方模型。|\n|[OpenClaw](https://github.com/openclaw/openclaw)|![GitHub Repo stars](https://badgen.net/github/stars/openclaw/openclaw)|\"龙虾\"——开源个人 AI 助理，本地运行并接入飞书 / 微信 / Telegram 等 IM，支持 Skills 与 MCP，配套 ClawHub 技能市场，是 2026 年最火的个人 Agent 之一。|\n|[Hermes Agent](https://github.com/NousResearch/hermes-agent)|![GitHub Repo stars](https://badgen.net/github/stars/NousResearch/hermes-agent)|Nous Research 出品的开源自我改进个人 Agent，跨 CLI / 桌面 / IM 运行，跨会话学习，可驱动真实终端与浏览器。|\n|[pi](https://github.com/earendil-works/pi)|![GitHub Repo stars](https://badgen.net/github/stars/earendil-works/pi)|极简可扩展的编码 Agent harness（\"primitives, not features\"），内核仅 Read/Write/Edit/Bash 四个工具，靠扩展与 Skills 定制，OpenClaw 即构建于其上。|\n|[Agent Skills（开放标准）](https://github.com/agentskills/agentskills)|![GitHub Repo stars](https://badgen.net/github/stars/agentskills/agentskills)|Anthropic 发起的 `SKILL.md` 开放标准，让 Agent 按需加载可复用技能，已被 Claude Code、Codex、Gemini CLI、Cursor、OpenCode、dsh 等 40+ 产品采纳。|\n\n\n## 前沿项目\n|名称|Stars|简介|备注|\n|---|---|---|---|\n|[:fire: AutoGPT](https://github.com/Significant-Gravitas/AutoGPT) |![GitHub Repo stars](https://badgen.net/github/stars/Significant-Gravitas/AutoGPT)|An experimental open-source attempt to make GPT-4 fully autonomous.|大名鼎鼎的 AutoGPT 项目, AI Agents 的早期尝试之一 (2023, 活跃度下降)|\n|[:fire: gpt-engineer](https://github.com/AntonOsika/gpt-engineer)|![GitHub Repo stars](https://badgen.net/github/stars/AntonOsika/gpt-engineer)|Specify what you want it to build, the AI asks for clarification, and then builds it.|全栈开发 Agent，用 GPT 编写整个项目代码！|\n|[:fire: AgentGPT](https://github.com/reworkd/AgentGPT) |![GitHub Repo stars](https://badgen.net/github/stars/reworkd/AgentGPT)|Assemble, configure, and deploy autonomous AI Agents in your browser.|在浏览器中部署运行 AI Agents.|\n|[:fire: MetaGPT](https://github.com/FoundationAgents/MetaGPT) |![GitHub Repo stars](https://badgen.net/github/stars/FoundationAgents/MetaGPT)|🌟 The Multi-Agent Framework: Given one line Requirement, return PRD, Design, Tasks, Repo. |MetaGPT, 多智能体框架，输入一句话的老板需求，输出用户故事 / 竞品分析 / 需求 / 数据结构 / APIs / 文件等|\n|[llama-hub,shopify agent](https://github.com/run-llama/llama-hub/blob/main/llama_hub/tools/notebooks/shopify.ipynb)|![GitHub Repo stars](https://badgen.net/github/stars/run-llama/llama-hub)|A customer support agent 🤖 that can interface with @Shopify’s ENTIRE GraphQL API Spec (\u003e50k lines!).| llama-hub 构建的客户支持 agent，能够与 @Shopify 的整个 GraphQL API规范(\u003e5万行!)交互|\n|[AGiXT (原 Agent-LLM)](https://github.com/Josh-XT/AGiXT)|![GitHub Repo stars](https://badgen.net/github/stars/Josh-XT/AGiXT)|An Artificial Intelligence Automation Platform. AI Instruction management from various providers, has an adaptive memory, and a versatile plugin system with many commands including web browsing.| 人工智能自动化平台 (项目已更名为 AGiXT)。https://agixt.com/|\n|[skyagi](https://github.com/litanlitudan/skyagi)|![GitHub Repo stars](https://badgen.net/github/stars/litanlitudan/skyagi)|SkyAGI implements the idea of Generative Agents and delivers a role-playing game that creates a very interesting user experience.| SkyAGI 实现了 \"生成式智能体 \"的理念，设计了一个角色扮演游戏，创造了非常有趣的用户体验。|\n|[generative_agents](https://github.com/joonspk-research/generative_agents)|![GitHub Repo stars](https://badgen.net/github/stars/joonspk-research/generative_agents)|Generative Agents: Interactive Simulacra of Human Behavior.| 斯坦福和谷歌的研究人员以《模拟人生》游戏为灵感，创建的 AI 智能体小镇；研究人员在模拟城镇中添加了 25 个生成式智能体 (Generative Agents)，这 25 个角色由 ChatGPT 和自定义代码控制，以高度逼真的行为独立地生活。在 ChatGPT 的支持下，每个人都有自己独特的身份、记忆和行为，并且可以独立交互，但他们都不会意识到自己是生活在模拟中。[中文介绍](https://www.oschina.net/news/253170/generative-agents-open-source)|\n|[developer](https://github.com/smol-ai/developer)|![GitHub Repo stars](https://badgen.net/github/stars/smol-ai/developer)|the first library to let you embed a developer agent in your own app!| 工程师智能体，给它一个产品规格，为你搭建一个完整的代码库，提供基本的模块，让您在自己的应用程序内拥有一个智能开发人员。|\n|[opencode](https://github.com/anomalyco/opencode)|![GitHub Repo stars](https://badgen.net/github/stars/anomalyco/opencode)|An AI coding agent built for the terminal.|SST 团队打造的终端编程 Agent，模型无关，支持多种 LLM 提供商，2025 年增长极快。|\n|[Aider](https://github.com/Aider-AI/aider)|![GitHub Repo stars](https://badgen.net/github/stars/Aider-AI/aider)|AI pair programming in your terminal.|终端 AI 结对编程工具，直接在本地 Git 仓库中编辑代码并自动提交，是 CLI 编程 Agent 的代表作。|\n|[Tura](https://github.com/Tura-AI/tura)|![GitHub Repo stars](https://badgen.net/github/stars/Tura-AI/tura)|A local-first open-source coding agent with terminal and graphical interfaces.|使用 Rust 编写的编程 Agent，提供 CLI、TUI 与 GUI，支持项目级规则、检查点、内置验证流程和公开基准测试。|\n|[goose](https://github.com/aaif-goose/goose)|![GitHub Repo stars](https://badgen.net/github/stars/aaif-goose/goose)|An open source, extensible AI agent that goes beyond code suggestions.|Block(Square 母公司)开源的本地可扩展 AI Agent，支持 MCP，可自动构建、执行和调试任务。|\n|[DeerFlow](https://github.com/bytedance/deer-flow)|![GitHub Repo stars](https://badgen.net/github/stars/bytedance/deer-flow)|A community-driven Deep Research framework.|字节跳动开源的深度研究 Agent 框架，基于 LangGraph，结合搜索、爬取与代码执行完成自动化研究报告。|\n|[Suna](https://github.com/kortix-ai/suna)|![GitHub Repo stars](https://badgen.net/github/stars/kortix-ai/suna)|Open source generalist AI agent.|开源通用型 AI Agent（对标 Manus），可自主浏览网页、执行命令、操作文件完成真实世界任务。|\n\n\n## Agents 开发平台\n|名称|Stars|简介|备注|\n|---|---|---|---|\n|[langchain](https://github.com/langchain-ai/langchain)|![GitHub Repo stars](https://badgen.net/github/stars/langchain-ai/langchain)|Building applications with LLMs through composability|开发的 ChatGPT 应用,构建基于 LLM 的 agents. 参见 [LangChain Agents 官方文档](https://docs.langchain.com/oss/python/langchain/agents) |\n|[AutoChain](https://github.com/Forethought-Technologies/AutoChain)|![GitHub Repo stars](https://badgen.net/github/stars/Forethought-Technologies/AutoChain)|AutoChain: Build lightweight, extensible, and testable LLM Agents.| AutoChain:构建轻量级、可扩展和可测试的LLM agents。|\n|[SuperAGI](https://github.com/TransformerOptimus/SuperAGI)|![GitHub Repo stars](https://badgen.net/github/stars/TransformerOptimus/SuperAGI)|\u003c⚡️\u003e SuperAGI - A dev-first open source autonomous AI agent framework. Enabling developers to build, manage \u0026 run useful autonomous agents quickly and reliably.|[官网](superagi.com/) 构建、管理和运行 AI Agents.|\n|[superagent](https://github.com/superagent-ai/superagent)|![GitHub Repo stars](https://badgen.net/github/stars/superagent-ai/superagent)|Superagent - Build, deploy, and manage LLM-powered agents.|[官网](https://docs.superagent.sh/) 开发人员能更轻松地构建、管理和部署智能体到生产中，包括内置内存、通过向量数据库检索文档、强大的工具、网络钩子、cron 任务等功能.|\n|[ai-town](https://github.com/a16z-infra/ai-town)|![GitHub Repo stars](https://badgen.net/github/stars/a16z-infra/ai-town)|A MIT-licensed, deployable starter kit for building and customizing your own version of AI town - a virtual town where AI characters live, chat and socialize.| 用于构建和定制你自己版本的人工智能小镇--一个人工智能角色生活、聊天和社交的虚拟小镇。|\n|[agent-protocol](https://github.com/agi-inc/agent-protocol)|![GitHub Repo stars](https://badgen.net/github/stars/agi-inc/agent-protocol)|Common interface for interacting with AI agents. The protocol is tech stack agnostic - you can use it with any framework for building agents.|[官网](www.agentprotocol.ai) AutoGPT, smol developer 等知名项目都在使用的\"智能体通讯协议\"--用于与智能体进行通信的通用接口。|\n|[autogen](https://github.com/microsoft/autogen)|![GitHub Repo stars](https://badgen.net/github/stars/microsoft/autogen)|AutoGen is a framework that enables development of LLM applications using multiple agents that can converse with each other to solve task.| AutoGen是一个框架，它允许使用多个 agents 来开发LLM应用程序，这些智能体可以相互交谈以解决任务。|\n|[agents](https://github.com/aiwaves-cn/agents)|![GitHub Repo stars](https://badgen.net/github/stars/aiwaves-cn/agents)|An Open-source Framework for Autonomous Language Agents.| Agents 是一个用于构建自主语言智能体的开源库/框架。|\n|[bisheng 毕昇](https://github.com/dataelement/bisheng)|![GitHub Repo stars](https://badgen.net/github/stars/dataelement/bisheng)|Bisheng is an open LLM devops platform for next generation AI applications. | 一款领先的开源大模型应用开发平台，赋能和加速大模型应用开发落地，帮助用户以最佳体验进入下一代应用开发模式。|\n|[Agently](https://github.com/Maplemx/Agently)|![GitHub Repo stars](https://badgen.net/github/stars/Maplemx/Agently)|🚀 A fast way to build LLM Agent based Application. | 面向应用开发者：Agently提供方便快速生成能力强大的Agent实例的能力，让开发者可以便捷地将这些实例与自己的业务代码相结合。|\n|[Dify](https://github.com/langgenius/dify)|![GitHub Repo stars](https://badgen.net/github/stars/langgenius/dify)|Production-ready platform for agentic workflow development.|生产级 LLM 应用与 Agent 工作流开发平台，提供可视化编排、RAG、工具调用与可观测性，社区规模巨大。|\n|[Agno](https://github.com/agno-agi/agno)|![GitHub Repo stars](https://badgen.net/github/stars/agno-agi/agno)|Full-stack framework for building Multi-Agent Systems with memory, knowledge and reasoning.|原 Phidata，高性能全栈多 Agent 框架，内置记忆、知识、推理与工具，主打轻量与运行速度。|\n|[Letta](https://github.com/letta-ai/letta)|![GitHub Repo stars](https://badgen.net/github/stars/letta-ai/letta)|Letta (formerly MemGPT) is a framework for creating stateful LLM agents.|原 MemGPT，专注于具备长期记忆的有状态 Agent，让智能体在跨会话中持续学习与保留上下文。|\n|[Aeon](https://github.com/aeonfun/aeon)|![GitHub Repo stars](https://badgen.net/github/stars/aeonfun/aeon)|The most autonomous agent framework - skills as Markdown files, scheduled on GitHub Actions with no approval loops, self-heals broken skills, and dispatches each skill to one of six model harnesses.|完全自主的 Agent 框架，技能即 Markdown 文件，通过 cron 定时在 GitHub Actions 上无人值守运行，可自我修复损坏的技能，支持 Claude、Grok、Codex、Pi、Vibe、Kimi 六种模型。|\n\n## Curated List\n|名称|Stars|简介|备注|\n|---|---|---|---|\n|[awesome-ai-agents](https://github.com/e2b-dev/awesome-ai-agents)|![GitHub Repo stars](https://badgen.net/github/stars/e2b-dev/awesome-ai-agents)|A list of AI autonomous agents. |基于 LLM 的 agents 精选资源.|\n|[awesome-llm-apps](https://github.com/Shubhamsaboo/awesome-llm-apps)|![GitHub Repo stars](https://badgen.net/github/stars/Shubhamsaboo/awesome-llm-apps)|Collection of awesome LLM apps with AI Agents and RAG.|大量带 AI Agent 与 RAG 的实战应用示例合集，覆盖多种模型与框架，实操性强。|\n|[awesome-mcp-servers](https://github.com/punkpeye/awesome-mcp-servers)|![GitHub Repo stars](https://badgen.net/github/stars/punkpeye/awesome-mcp-servers)|A collection of MCP servers.|MCP(Model Context Protocol)服务器精选合集，是为 Agent 接入外部工具/数据源的重要资源导航。|\n\n## Star History\n\n[![Star History Chart](https://api.star-history.com/svg?repos=yzfly/Awesome-AGI-Agents\u0026type=Date)](https://star-history.com/#yzfly/Awesome-AGI-Agents\u0026Date)\n\n## License\n\n本项目采用 [CC BY-NC 4.0](LICENSE) 协议授权 (署名-非商业性使用)。\n\n---\n\n作者：云中江树，微信公众号: 云中江树\n","projects_url":"https://awesome.ecosyste.ms/api/v1/lists/yzfly%2Fawesome-agi-agents/projects"}