Agent Architectures

9 pages in this section.

The Three Core Agent Architecture Patterns, Compared

Introduces ReAct, Plan-and-Execute, and multi-agent handoff as the field's three dominant patterns, and how they differ in control flow, cost, and failure modes.

Python Agent Architectures Explainer React Pattern

Agent Architectures Basics

A first comparison of a ReAct-style loop against a simple planned sequence of steps, with minimal Python sketches you can reason about before choosing a framework.

Python Agent Architectures Basics React Pattern

ReAct: Reasoning and Acting in an Interleaved Loop

Explains the reason-then-act-then-observe cycle that gives ReAct its name, why interleaving helps tool use, and what you must bound for production reliability.

Python Agent Architectures Explainer React Pattern

Plan-and-Execute: Separating Planning from Execution

Explains why generating a full plan upfront reduces wasted tool calls on complex tasks, how execution and replan hooks work, and when phased control beats pure ReAct.

Python Agent Architectures Explainer Plan And Execute

Multi-Agent Handoff: Specialist Agents Passing Work to Each Other

Python Agent Architectures Explainer Multi Agent

When to Choose ReAct vs Plan-and-Execute vs Multi-Agent

A decision cheatsheet matching task complexity, predictability, privilege needs, and cost constraints to ReAct, Plan-and-Execute, multi-agent, or a hybrid.

Python Agent Architectures List Cheatsheet

Hybrid Architectures: Combining Planning, Reaction, and Delegation

Python Agent Architectures Hybrid Planning

Common Architecture Anti-Patterns in Early Agent Projects

Python Agent Architectures List Anti Patterns

Agent Architectures Best Practices

Ten guidelines for matching architecture to task shape, bounding loops, scoping tools, and documenting control flow before writing agent orchestration code.

Python Agent Architectures Best Practices List

Agent Architectures

Search Documentation

Agent Architectures