AI Agents in Practice
Designing Assistants That Plan, Use Tools and Stay Safe
- Pages
- 81
- Chapters
- 17
- Level
- Advanced
- Format
- E-book
About this book
An AI agent does more than answer: it decides which steps to take, calls tools such as search, databases or email, looks at the results and keeps going until the job is done. That power brings new risks, from runaway loops to actions nobody approved.
AI Agents in Practice is a design guide for building agents that are useful and safe. It explains the agent loop, how to design tools a model can use correctly, planning and task decomposition, short- and long-term memory, and how to put permissions, confirmations and human oversight in the right places. It also covers testing and monitoring, which are harder for agents than for ordinary software.
The last chapter presents proven design patterns, and, just as valuable, the signs that a simpler workflow would serve you better than an agent.
What you will learn
- What distinguishes an agent from a chatbot or a fixed workflow
- How to design tools and tool descriptions that models use correctly
- The agent loop, stopping conditions and error handling
- Planning, decomposition, memory and state management
- Guardrails: permissions, confirmations, sandboxing and human review
- Testing, tracing and monitoring agents in production
- When a workflow or state machine beats a free-roaming agent
- How to threat-model an agent and calculate its cost and reliability
- Multi-agent systems, computer-use and coding agents, and a complete case study
Table of contents
- IntroductionIntroduction
- Chapter 1What Makes Something an Agent
- Chapter 2Tools: Giving a Model Hands
- Chapter 3The Agent Loop
- Chapter 4Planning and Decomposition
- Chapter 5Memory and State
- Chapter 6Workflows and State Machines
- Chapter 7Multi-Agent Systems
- Chapter 8Agents That Use Computers
- Chapter 9Guardrails and Human Oversight
- Chapter 10Threat Modelling an Agent
- Chapter 11Testing and Monitoring Agents
- Chapter 12Cost, Latency and Reliability
- Chapter 13Case Study: An Invoice-Chasing Assistant
- Chapter 14Coding Agents
- Chapter 15Introducing Agents to a Team
- Chapter 16Choosing Models, Frameworks and Protocols
- Chapter 17Patterns, and When Not to Build an Agent
- AppendixAgent Design Review Worksheet
- AppendixExercises
- GlossaryGlossary
Who this book is for
Software engineers, technical leads and product builders designing AI features that take actions. Familiarity with language models and basic programming is assumed.
Details
| Format | PDF + web edition |
|---|---|
| Length | 81 pages (approx.) |
| Chapters | 17 |
| Level | Advanced |
| Language | English |
| Topic | AI Engineering |
| Author | Latent Library Editorial |
Keep reading