Technical book recommendations specifically on AI agent planning, tool calling & orchestration
I’m a software engineer currently working on production AI agent/workflow systems. I already have a strong software/system-design background, so I’m not looking for general system-design books, basic ML textbooks, or broad introductions to LLMs.
I want to go deep specifically on agentic AI architecture, especially:
- Agent planning and task decomposition
- ReAct / reflection / replanning loops
- Tool/function calling and tool selection
- Designing good tools/interfaces for agents
- Planner–executor architectures
- Agent loops: plan → act → observe → replan
- Workflow vs autonomous-agent design
- Orchestration and routing
- Supervisor/worker and coordinator/delegator patterns
- Multi-agent communication and delegation
- State, memory and context in agent workflows
- MCP / A2A where relevant
- Evaluation and debugging of multi-step agent behavior
- Failure recovery, stopping conditions and guardrails
I care much more about depth and syllabus relevance than popularity or Amazon ratings. A newer or less-known book is completely fine if it genuinely covers these topics well.
Some books I’ve come across:
- Agentic Design Patterns — Antonio Gulli
- AI Agents in Action, 2nd Edition — Micheal Lanham
- AI Agents: The Definitive Guide — Nicole Koenigstein
- Building the Agentic Enterprise — Jaroslav Pantsjoha
- Building Agentic AI Systems — Anjanava Biswas & Wrick Talukdar
- Systems Thinking for Agentic AI
Has anyone actually read any of these?
If you wanted to become really strong at designing production agent planning/tool-use/orchestration systems, which 1–3 books would you choose?
Also very interested in books I haven't listed. Papers/blogs/courses are welcome too if you think books haven't caught up with the field yet.