build-your-gemini-code-from-scratch

# Agent Learning Project

**[中文版本 / Chinese Version](README_zh.md)**

## 🚀 Experience Quick Star CLI in Action!

![Quick Star CLI](resources/images/quick_star_cli.png)

**✨ Try Quick Star CLI now!** An intelligent AI agent with elegant command-line interface, real-time streaming responses, and powerful code generation capabilities.

## 🎮 See What's Possible: Snake Game Created by AI!
![Snake Game](resources/images/snake_game.jpg)

**🎯 Play the snake game!** This fully functional game was created entirely through natural language conversations with Quick Star CLI, showcasing:
- 🧠 **Intelligent Code Generation** - Complete game logic from simple descriptions
- 🔧 **Real-time Debugging** - Iterate and improve code instantly  
- 📁 **Smart File Management** - Handle complex project structures
- 🎨 **Interactive Applications** - Create engaging user experiences

**Ready to build something amazing?** Follow the installation guide below and start creating with Quick Star CLI!

---


This project demonstrates the progressive development of AI agents, from basic tool calling to advanced streaming agents with history control. Each chapter builds upon the previous one, showing incremental improvements and new features.

## Project Structure

```
├── .env                          # Environment configuration (shared by all chapters)
├── .env.example                  # Environment template
├── requirements.txt              # Python dependencies
├── chapter1_tool_call_api/       # Basic tool calling examples (Native Function Call & XML Tool Call)
├── chapter2_ReAct_agent/         # Basic ReAct agent implementation
├── chapter3_stream_agent/        # Streaming agent with real-time responses
├── chapter4_history_control/     # Advanced agent with conversation history management
├── chapter5_smart_context/       # Smart context management with intelligent cropping
├── chapter6_to_do_write/         # Task management with TodoWrite tool
└── chapter7_sub_agent/           # Sub-agent