English | 中文
Preface
Since the end of 2022, Large Language Models (LLMs) represented by ChatGPT have swept across the world like a technological tsunami, completely transforming how we interact with artificial intelligence. The powerful natural language understanding and generation capabilities of LLMs have shown us a glimpse of the path toward Artificial General Intelligence (AGI). However, as the initial amazement settled, developers began exploring the next frontier: how to make AI not just a "question-answering" tool, but an "actor" capable of autonomous planning, tool invocation, and solving complex problems?
The answer is Agents.
If 2024 was the inaugural year of the "battle of a hundred models," then 2025 has undoubtedly ushered in the "Year of Agents." We see that the technological focus is shifting from training larger and more powerful foundation models to building smarter and more efficient agent applications. Individual agents can already handle tasks in specific domains, while Multi-Agent Systems (MAS), where multiple agents collaborate through division of labor, cooperation, and even debate to accomplish grand goals, are viewed as the key to unlocking the full potential of LLMs and solving complex real-world problems.
However, there is an obvious gap in the current ecosystem: on one hand, there is a dizzying array of Agent frameworks and applications emerging continuously; on the other hand, there is an extreme scarcity of systematic knowledge. Most tutorials focus on API calls for specific frameworks, leaving learners "knowing how but not knowing why," still feeling powerless when facing complex requirements. We lack a practical guide that can penetrate framework appearances, start from first principles, and systematically explain agent design, construction, and collaboration.
In view of this, we launched the Hello-Agents project, hoping to provide the community with a guide for building agent systems from scratch, balancing theory and practice. We will not only lead you to appreciate the most cutting-edge technologies in the agent field but also guide you to delve into their core architecture, understand their classic paradigms, and ultimately build your own multi-agent applications with your own hands.
We believe that the best way to learn is through hands-on practice. We hope this tutorial can become your starting point for exploring the world of agents, enabling you to transform from a "user" of large language models to a "builder" of agent systems.
Suggestions for Readers
Welcome, future intelligent system builder! Before embarking on this exciting journey, please allow us to give you some small suggestions.
Before reading this project, we hope you:
Have basic Python programming skills.
Have a basic conceptual understanding of large language models (for example, know how to obtain LLM APIs).
Rest assured, you don't need a deep background in algorithms or model training; the project focuses on application and construction.
This project is divided into five parts, covering basics to practice, progressing step by step, layer by layer:
Part One (Fundamentals): We will lay the foundation of core knowledge about artificial intelligence and LLMs, giving you a macro understanding of the background of agent emergence.
Part Two (Single Agent): This is where your hands-on practice begins. We will guide you to build a fully functional single agent from scratch, deeply understanding its internal "mental" structure.
Part Three (Advanced): Here, your agent will "learn" to think, possess memory and tools, and master communication protocols between agents, ultimately completing the evaluation closed loop.
Part Four (Practice): This is where the core value of the project lies. You will integrate all learned knowledge through a series of carefully designed comprehensive cases, tempering true gold in practice.
Part Five (Outlook): The end of the journey is a new beginning. You will personally create your "graduation project," drawing a perfect conclusion to your learning journey.
"What is learned on paper is superficial; to truly understand, one must practice." To achieve the best learning effect, we provide all supporting code in the project's code folder. We strongly recommend combining theory with practice. Please be sure to personally run, debug, and even modify every piece of code provided in the project. We encourage you to apply what you've learned to real scenarios that interest you—this is the ultimate purpose of learning.
Finally, as an open-source project, we warmly welcome your participation and contribution. When you encounter problems, you can ask questions in our community; when you have new ideas or discoveries, you are also welcome to join the project's co-construction at any time.
Thank you for choosing to read Hello-Agents. We wish you happy learning and unlimited exploration!