Reference Book

Agents Reference

A comprehensive reference guide to AI agents, agentic workflows, and orchestration patterns. Designed for practitioners building intelligent systems.

What are AI Agents?

Definition, capabilities, and the agentic loop

Read the overview

How to Use This Reference

Core Terms

Foundational definitions

Patterns

Proven architectures

Workflows

Esy implementations

Core Terms

Foundational definitions for understanding AI agent systems

Agentic Workflow

Multi-step AI task automation

Agentic Loop

The perceive-reason-act-evaluate cycle

Tool Use

How agents interact with external systems

Memory

Short-term and long-term context retention

Planning

Goal decomposition and task sequencing

Orchestration

Coordinating multiple agents or steps

Human-in-the-Loop

Human oversight in autonomous systems

Architecture Patterns

Proven approaches for structuring agent systems

Planner-Executor

Separate planning from execution

Router-Supervisor

Dynamic task routing with oversight

Reflect-Revise

Self-evaluation and iterative improvement

Esy Workflows

Implementation examples showing patterns in practice

Gated Agent Pipeline

Quality gates between agent steps

Research Agent Workflow

End-to-end research automation

Esy Implementation Track: Where relevant, entries include practical implementation examples from Esy's research workflows. These demonstrate how abstract concepts translate to working systems.