Build your first agent
Learn what AI agents are, work with models and prompts, and set up your Python environment. Build your first agent, design tools, and connect them through MCP.
Agent loops · Tool design · MCP
← All booksBuilding AI Agents
A practical guide with Python
Building AI Agents
from Scratch
A Hands-On Guide with Python
Turn an idea into an agent that can act. Learn the tools, memory, coordination, and evaluation behind useful AI systems — one chapter at a time.
One purchase in USD. PDF + Python code.
READ. BUILD. EXPERIMENT.
YOUR NEXT BUILD
STARTS HERE.
STARTS HERE.
21Chapters
4Appendices
PythonCompanion examples
Yours to keepRead & build offline
TAKE A LOOK INSIDE
Understand the pieces.
Then put them to work.
For Python developers who want to move from talking to a model to building systems around one.
Add context and coordinate agents
Manage memory and state, plan actions, and retrieve documents with RAG. Learn when multiple agents help, how they communicate, and how to build coordinated systems.
Memory · Planning · Multi-agent systems
Evaluate, safeguard, and deploy
Evaluate and test agent behavior, handle failures and safety boundaries, and prepare agents for production deployment.
Evaluation · Recovery · Deployment
Extend what your agents can do
Stream responses, explore fine-tuning, work with multimodal inputs, query databases with SQL, and build voice interfaces for agents.
SQL · Voice · Multimodal agents
YOUR DIGITAL EDITION
Everything in
your download.
A single ZIP with your purchased files. Save it, unpack it, and start reading.
Read a sampleThe complete ebook
Selectable text, a linked table of contents, and clear code typography.
Companion Python code
Examples and supporting files organized by chapter.
BEFORE YOU BEGIN
A few good
questions.
Is this book right for me?
The book is written in English for readers with basic Python knowledge. It introduces the building blocks of AI agents and progresses to more advanced systems. Read the sample to see whether the approach fits you.
Do I need paid AI tools to follow along?
Some examples run locally. Others require optional software, provider accounts, or paid API usage. Those services are not included in the ebook price. The code includes both runnable projects and teaching snippets.
How do I get the book after paying?
After your payment is confirmed, a download button appears on the purchase confirmation page. We also email your download link. You do not need to create an account.
Is this a subscription?
No. Each book is a one-time purchase. Download your files and keep them for offline use.
Can I get a refund?
Yes. Email benky.tide@gmail.com within 7 calendar days of purchase for a full refund, including if you have downloaded the files. See the refund policy for the request process.
How may I use the files?
The ebook is for personal use; the companion source code has its own MIT license.
Who publishes these books, and how can I get help?
AI Craft Mastery is an educational publishing website by Yanfei Li, an independent author based in Knoxville, Tennessee, United States. For book questions, purchase help or refunds, email benky.tide@gmail.com. Read our purchase and privacy information.
YOUR NEXT CHAPTER
Start building.
Building AI Agents from Scratch
PDF + Python code · One-time purchase