How bojieli/ai-agent-book works
This repository is an AI-agent engineering book accompanied by runnable experiments, not a single deployed agent. The main user journey is to read a chapter and follow its linked experiment to explore the corresponding technique. The README establishes the ten-chapter topics, book locations, and companion-experiment structure; the tree shows their implementation areas. Sampled source confirms selected experiment internals, but not wiring across the unsampled projects, so the graph avoids inferring those connections.
Main entry points: book Chinese book · book-en English book · book-es Other translations · chapter1 Agent introduction
Components 15
Book editions
- Chinese book
book - English book
book-en - Other translations
book-es
Agent foundations
- Agent introduction
chapter1 - Context engineering
chapter2 - Provider plumbing
agentbook/providers
Memory and tools
- Memory and knowledge
chapter3 - Agent tools
chapter4 - Coding agents
chapter5
Interaction and evaluation
- Interaction
chapter6 - Agent evaluation
chapter7
Training and evolution
- Model post-training
chapter8 - Agent evolution
chapter9 - Production agents
chapter10
Other components
- Reader
Connections 13
- Reader → Chinese book: reads
- Reader → English book: reads
- Reader → Other translations: reads
- Chinese book → Agent introduction: links experiments
- Chinese book → Context engineering: links experiments
- Chinese book → Memory and knowledge: links experiments
- Chinese book → Agent tools: links experiments
- Chinese book → Coding agents: links experiments
- Chinese book → Interaction: links experiments
- Chinese book → Agent evaluation: links experiments
- Chinese book → Model post-training: links experiments
- Chinese book → Agent evolution: links experiments
- Chinese book → Production agents: links experiments