AutoAgent

<a name="readme-top"></a>

<div align="center"> <img src="./assets/AutoAgent_logo.svg" alt="Logo" width="200"> <h1 align="center">AutoAgent: Fully-Automated & Zero-Code</br> LLM Agent Framework </h1> </div> <div align="center"> <a href="https://autoagent-ai.github.io"><img src="https://img.shields.io/badge/Project-Page-blue?style=for-the-badge&color=FFE165&logo=homepage&logoColor=white" alt="Credits"></a> <a href="https://join.slack.com/t/metachain-workspace/shared_invite/zt-2zibtmutw-v7xOJObBf9jE2w3x7nctFQ"><img src="https://img.shields.io/badge/Slack-Join%20Us-red?logo=slack&logoColor=white&style=for-the-badge" alt="Join our Slack community"></a> <a href="https://discord.gg/jQJdXyDB"><img src="https://img.shields.io/badge/Discord-Join%20Us-purple?logo=discord&logoColor=white&style=for-the-badge" alt="Join our Discord community"></a> <!-- <a href="https://github.com/HKUDS/AutoAgent/blob/main/assets/autoagent-wechat.jpg"><img src="https://img.shields.io/badge/Wechat-Join%20Us-green?logo=wechat&logoColor=white&style=for-the-badge" alt="Join our Wechat community"></a> --> <a href="./Communication.md"><img src="https://img.shields.io/badge/💬Feishu-Group-07c160?style=for-the-badge&logoColor=white&labelColor=1a1a2e"></a> <a href="./Communication.md"><img src="https://img.shields.io/badge/WeChat-Group-07c160?style=for-the-badge&logo=wechat&logoColor=white&labelColor=1a1a2e"></a> <br/> <a href="https://autoagent-ai.github.io/docs"><img src="https://img.shields.io/badge/Documentation-000?logo=googledocs&logoColor=FFE165&style=for-the-badge" alt="Check out the documentation"></a> <a href="https://arxiv.org/abs/2502.05957"><img src="https://img.shields.io/badge/Paper%20on%20Arxiv-000?logoColor=FFE165&logo=arxiv&style=for-the-badge" alt="Paper"></a> <a href="https://gaia-benchmark-leaderboard.hf.space/"><img src="https://img.shields.io/badge/GAIA%20Benchmark-000?logoColor=FFE165&logo=huggingface&style=for-the-badge" alt="Evaluation Benchmark Score"></a> <hr> </div> <div align="center"> <a href="https://trendshift.io/repositories/13954" target="_blank"><img src="https://trendshift.io/api/badge/repositories/13954" alt="HKUDS%2FAutoAgent | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a> </div>

Welcome to AutoAgent! AutoAgent is a Fully-Automated and highly Self-Developing framework that enables users to create and deploy LLM agents through Natural Language Alone.

✨Key Features of AutoAgent

• 💬 Natural Language-Driven Agent Building </br>Automatically constructs and orchestrates collaborative agent systems purely through natural dialogue, eliminating the need for manual coding or technical configuration.

• 🚀 Zero-Code Framework </br>Democratizes AI development by allowing anyone, regardless of coding experience, to create and customize their own agents, tools, and workflows using natural language alone.

• ⚡ Self-Managing Workflow Generation </br>Dynamically creates, optimizes and adapts agent workflows based on high-level task descriptions, even when users cannot fully specify implementation details.

• 🔧 Intelligent Resource Orchestration </br>Enables controlled code generation for creating tools, agents, and workflows through iterative self-improvement, supporting both single agent creation and multi-agent workflow generation.

• 🎯 Self-Play Agent Customization </br>Enables controlled code generation for creating tools, agents, and workflows through iterative self-improvement, supporting both single agent creation and multi-agent workflow generation.

🚀 Unlock the Future of LLM Agents. Try 🔥AutoAgent🔥 Now!

<div align="center"> <!-- <img src="./assets/AutoAgentnew-intro.pdf" alt="Logo" width="100%"> --> <figure> <img src="./assets/autoagent-intro.svg" alt="Logo" style="max-width: 100%; height: auto;"> <figcaption><em>Quick Overview of AutoAgent.</em></figcaption> </figure> </div>

🔥 News

<div class="scrollable"> <ul> <li><strong>[2025, Feb 17]</strong>: &nbsp;🎉🎉We've updated and released AutoAgent v0.2.0 (formerly known as MetaChain). Detailed changes include: 1) fix the bug of different LLM providers from issues; 2) add automatic installation of AutoAgent in the container environment according to issues; 3) add more easy-to-use commands for the CLI mode. 4) Rename the project to AutoAgent for better understanding.</li> <li><strong>[2025, Feb 10]</strong>: &nbsp;🎉🎉We've released <b>MetaChain!</b>, including framework, evaluation codes and CLI mode! Check our <a href="https://arxiv.org/abs/2502.05957">paper</a> for more details.</li> </ul> </div> <span id='table-of-contents'/>

📑 Table of Contents

• <a href='#features'>✨ Features</a>
• <a href='#news'>🔥 News</a>
• <a href='#how-to-use'>🔍 How to Use AutoAgent</a>
  • <a href='#user-mode'>1. user mode (Deep Research Agents)</a>
  • <a href='#agent-editor'>2. agent editor (Agent Creation without Workflow)</a>
  • <a href='#workflow-editor'>3. workflow editor (Agent Creation with Workflow)</a>
• <a href='#quick-start'>⚡ Quick Start</a>
  • <a href='#installation'>Installation</a>
  • <a href='#api-keys-setup'>API Keys Setup</a>
  • <a href='#start-with-cli-mode'>Start with CLI Mode</a>
• <a href='#todo'>☑️ Todo List</a>
• <a href='#reproduce'>🔬 How To Reproduce the Results in the Paper</a>
• <a href='#documentation'>📖 Documentation</a>
• <a href='#community'>🤝 Join the Community</a>
• <a href='#acknowledgements'>🙏 Acknowledgements</a>
• <a href='#cite'>🌟 Cite</a>

<span id='how-to-use'/>

🔍 How to Use AutoAgent

<span id='user-mode'/>

1. user mode (Deep Research Agents)

AutoAgent features a ready-to-use multi-agent system accessible through user mode on the start page. This system serves as a comprehensive AI research assistant designed for information retrieval, complex analytical tasks, and comprehensive report generation.

• 🚀 High Performance: Matches Deep Research using Claude 3.5 rather than OpenAI's o3 model.
• 🔄 Model Flexibility: Compatible with any LLM (including Deepseek-R1, Grok, Gemini, etc.)
• 💰 Cost-Effective: Open-source alternative to Deep Research's $200/month subscription
• 🎯 User-Friendly: Easy-to-deploy CLI interface for seamless interaction
• 📁 File Support: Handles file uploads for enhanced data interaction

<div align="center"> <video width="80%" controls> <source src="./assets/video_v1_compressed.mp4" type="video/mp4"> </video> <p><em>🎥 Deep Research (aka User Mode)</em></p> </div> <span id='agent-editor'/>

2. agent editor (Agent Creation without Workflow)

The most distinctive feature of AutoAgent is its natural language customization capability. Unlike other agent frameworks, AutoAgent allows you to create tools, agents, and workflows using natural language alone. Simply choose agent editor or workflow editor mode to start your journey of building agents through conversations.

You can use agent editor as shown in the following figure.

<table> <tr align="center"> <td width="33%"> <img src="./assets/agent_editor/1-requirement.png" alt="requirement" width="100%"/> <br> <em>Input what kind of agent you want to create.</em> </td> <td width="33%"> <img src="./assets/agent_editor/2-profiling.png" alt="profiling" width="100%"/> <br> <em>Automated agent profiling.</em> </td> <td width="33%"> <img src="./assets/agent_editor/3-profiles.png" alt="profiles" width="100%"/> <br> <em>Output the agent profiles.</em> </td> </tr> </table> <table> <tr align="center"> <td width="33%"> <img src="./assets/agent_editor/4-tools.png" alt="tools" width="100%"/> <br> <em>Create the desired tools.</em> </td> <td width="33%"> <img src="./assets/agent_editor/5-task.png" alt="task" width="100%"/> <br> <em>Input what do you want to complete with the agent. (Optional)</em> </td> <td width="33%"> <img src="./assets/agent_editor/6-output-next.png" alt="output" width="100%"/> <br> <em>Create the desired agent(s) and go to the next step.</em> </td> </tr> </table> <span id='workflow-editor'/>

3. workflow editor (Agent Creation with Workflow)

You can also create the agent workflows using natural language description with the workflow editor mode, as shown in the following figure. (Tips: this mode does not support tool creation temporarily.)

<table> <tr align="center"> <td width="33%"> <img src="./assets/workflow_editor/1-requirement.png" alt="requirement" width="100%"/> <br> <em>Input what kind of workflow you want to create.</em> </td> <td width="33%"> <img src="./assets/workflow_editor/2-profiling.png" alt="profiling" width="100%"/> <br> <em>Automated workflow profiling.</em> </td> <td width="33%"> <img src="./assets/workflow_editor/3-profiles.png" alt="profiles" width="100%"/> <br> <em>Output the workflow profiles.</em> </td> </tr> </table> <table> <tr align="center"> <td width="33%"> <img src="./assets/workflow_editor/4-task.png" alt="task" width="66%"/> <br> <em>Input what do you want to complete with the workflow. (Optional)</em> </td> <td width="33%"> <img src="./assets/workflow_editor/5-output-next.png" alt="output" width="66%"/> <br> <em>Create the desired workflow(s) and go to the next step.</em> </td> </tr> </table> <span id='quick-start'/>

⚡ Quick Start

<span id='installation'/>

Installation

AutoAgent Installation

git clone https://github.com/HKUDS/AutoAgent.git
cd AutoAgent
pip install -e .

Docker Installation

We use Docker to containerize the agent-interactive environment. So please install Docker first. You don't ne