AI Briefingai-agent-book: Agent = LLM + Context + Tools, with 103 experiments
bojieli/ai-agent-book
About the project
This open-source book is structured into 10 chapters, centered on the core formula for AI agent design: 'Agent = LLM + Context + Tools'. It goes beyond simple conceptual explanations by providing 103 experiment codes matched to each chapter, allowing readers to directly verify the theory.
It broadly covers the latest trends, including context engineering, RAG, MCP protocol, coding agents, multimodal interaction, model post-training (SFT/RL), and multi-agent collaboration. Notably, version 2.0 integrates asynchronous interaction and multimodal agent content, adding a new chapter titled 'Expanding the Observation and Action Space'.
Translated into a total of 14 languages, including English, Korean, and Japanese, based on the original Chinese text, it is available for free download in PDF and EPUB formats. It supports Python 3.11~3.13 environments, allowing users to install per-chapter dependencies and run experiments via uv or pip.
bojieli/ai-agent-book
《深入理解 AI Agent：设计原理与工程实践》（李博杰 著）开源主仓库：全书正文、编译版 PDF 与按章配套代码
Python
This introduction was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report errors, attribution issues, or removal requests via Contact.