Agentic Design Patterns

📖 🔗 👉 Agentic Design Patterns: A Hands-On Guide to Building Intelligent Systems
By Antonio Gulli

EPUB

⬇️ 💾 👉 Download the EPUB file

About This Book

This book explores the fundamental design patterns needed to build intelligent AI agents. From prompt chaining and tool use to multi-agent collaboration and safety patterns, it provides practical, hands-on guidance for developers building the next generation of AI applications.

📚 Book Structure

• 21 Chapters covering core agentic design patterns
• 6 Appendices with advanced techniques and frameworks
• 400+ pages of comprehensive content
• Hands-on code examples in Python using LangChain, CrewAI, and Google ADK
• Real-world applications and use cases

🏗️ Repository Structure

├── agentic-design-patterns.epub  # Final EPUB file
├── agentic-design-patterns.md    # Main book content (Markdown)
├── markdown/                     # Individual chapter files downloaded from Google Docs
│   ├── chapter_*.md             # Chapter files
│   ├── appendix_*.md            # Appendix files
│   └── table_of_contents_epub.md # EPUB table of contents
├── scripts/                      # Build and utility scripts
│   ├── create_epub_with_toc.py  # Main EPUB builder
│   └── ...                      # Other utility scripts  
├── images/                       # Book images and diagrams
│   ├── cover.png                # Book cover
│   └── *.png                    # Chapter illustrations
├── styles/                       # EPUB styling
│   └── epub-styles.css          # CSS for EPUB formatting
├── backups/                      # Backup files
├── epub/                         # EPUB files
└── docs/                        # Documentation

🛠️ Building the EPUB

Prerequisites

• Python 3.7+
• Pandoc for EPUB conversion
• Required Python packages (see individual scripts)

Quick Build

# Build EPUB with cover image, table of contents, and navigation
python3 scripts/create_epub_with_toc.py

This creates a complete EPUB with:

• ✅ Cover image on first page
• ✅ Embedded table of contents
• ✅ Internal chapter navigation
• ✅ EPUB metadata cover
• ✅ Custom CSS styling
• ✅ Proper EPUB3 structure

Generated Output

The EPUB will be created at:

epub/agentic-design-patterns-complete.epub

📖 Chapters Overview

Foundation Patterns (Part 1)

1. Prompt Chaining - Sequential LLM interactions
2. Routing - Dynamic task delegation
3. Parallelization - Concurrent processing
4. Reflection - Self-correction mechanisms
5. Tool Use - External API integration
6. Planning - Multi-step strategy formulation
7. Multi-Agent Collaboration - Distributed intelligence

Advanced Patterns (Part 2)

8. Memory Management - State persistence
9. Learning and Adaptation - Continuous improvement
10. Model Context Protocol - Standardized communication
11. Goal Setting and Monitoring - Objective tracking

Integration Patterns (Part 3)

12. Exception Handling - Error recovery
13. Human-in-the-Loop - Interactive workflows
14. Knowledge Retrieval (RAG) - Information augmentation

Production Patterns (Part 4)

15. Inter-Agent Communication - Agent-to-agent protocols
16. Resource-Aware Optimization - Efficient resource usage
17. Reasoning Techniques - Advanced problem solving
18. Guardrails/Safety - Safe AI operations
19. Evaluation and Monitoring - Performance assessment
20. Prioritization - Task management
21. Exploration and Discovery - Adaptive learning

🔧 Development Scripts

• create_epub_with_toc.py - Main EPUB builder with cover and navigation
• fix_headings.py - Heading level normalization
• create_epub_with_images.py - EPUB with embedded images
• gdocs_to_markdown.py - Google Docs conversion
• Various validation and testing scripts

📝 Content Guidelines

• All chapter content uses consistent heading hierarchy
• Code examples include proper attribution and licenses
• Images are optimized for EPUB compatibility
• Cross-references use proper markdown linking

🚀 Getting Started

1. Clone this repository
2. Install Pandoc: brew install pandoc (macOS) or follow installation guide
3. Build EPUB: python3 scripts/create_epub_with_toc.py
4. Open EPUB: Double-click the generated file or use your preferred e-reader

📱 Compatibility

The generated EPUB works with:

• macOS Books app
• Moon+ Reader app
• ElevenReader app for audio books
• Kindle (via Calibre conversion)
• Adobe Digital Editions
• Calibre e-book management
• Most modern e-readers

🤝 Contributing

This book represents cutting-edge knowledge in agentic AI systems. While the main content is authored by Antonio Gulli, the build system and scripts welcome improvements.

📄 License

All rights belong to Antonio Gulli.