How zhangdaxie/ai-agent-book works
This repository is the companion codebase for a ten-chapter AI-agent engineering book, with experiments illustrating its concepts. The main user journey is to study the book’s progression from agent fundamentals through production practice and explore the corresponding chapter experiments. The README documents chapter topics and experiment counts; source excerpts show selected implementation examples, not a unified application runtime. Chapter nodes therefore represent documented subject areas, while the shared provider package is included as an internal support layer.
Main entry points: book Book manuscript · chapter7 Agent evaluation · chapter1 Agent basics · chapter2 Context engineering
Components 13
Agent foundations
- Book manuscript
book - Agent basics
chapter1 - Context engineering
chapter2 - Memory and knowledge
chapter3
Agent engineering
- Agent tools
chapter4 - Coding agents
chapter5 - Agent interaction
chapter6
Evaluation and learning
- Agent evaluation
chapter7 - Model post-training
chapter8 - Agent evolution
chapter9
Advanced agents
- Multi-agent systems
chapter10 - Model providers
agentbook/providers
Other components
- Reader
Connections 12
- Reader → Book manuscript: reads
README.md - Book manuscript → Agent basics: introduces
README.md - Book manuscript → Context engineering: covers
README.md - Book manuscript → Memory and knowledge: covers
README.md - Book manuscript → Agent tools: covers
README.md - Book manuscript → Coding agents: covers
README.md - Book manuscript → Agent interaction: covers
README.md - Book manuscript → Agent evaluation: covers
README.md - Book manuscript → Model post-training: covers
README.md - Book manuscript → Agent evolution: covers
README.md - Book manuscript → Multi-agent systems: covers
README.md - Agent evaluation → Model providers: uses