Agents

Your starting point for understanding agents, learning how they work, and finding tools in the directory.

Special reports

Start with our two-part editorial — concepts and stack in part 1, tools and guardrails in part 2.

Agents, part 1

ML, models, prompts, agents — a simple map of how the pieces connect.

Agents, part 2

Tools, memory, loops, and guardrails — how agents turn text into action.

What are AI agents?

AI Agents are autonomous or semi-autonomous software entities designed to perceive their environment and take actions to achieve specific goals. Unlike a single chat reply, agents often chain tool calls, maintain state across steps, and pursue an outcome with less hand-holding. Full definition in Dictionary · More on Learn

Why they matter now

The shift from “answer my question” to “do this work” is what drives today's agent buzz. Agents show up in coding assistants, sales outreach, research, operations, and customer workflows — often combining a language model with tools, memory, and guardrails. Knowing the landscape helps you pick the right layer: framework, builder, or ready-made worker.

Choose your path

Three ways the directory organizes agent products — pick the lane that matches you.

Agentic AI

Teams using autonomous workers

Goal-driven products — Lindy, Relevance, Relay, GenSpark, and similar.

I want an agent that plans and executes multi-step work.

Popular in Agentic AI

Agents family — all three lanes in one view.

Agent terms

Quick excerpts — see the full Dictionary for every term.

  • Agents

    AI Agents are autonomous or semi-autonomous software entities designed to perceive their environment and take actions to achieve specific goals.

  • Agent Harness

    An agent harness is the runtime wrapped around a language model that turns next-token prediction into an agent that can act.

  • MCP

    Model Context Protocol (MCP) is a framework developed by Anthropic that enables AI models to access external tools and share memory across different applications.

  • RAG

    Retrieval Augmented Generation (RAG) is an advanced AI architecture that combines large language models with information retrieval systems to generate more accurate and contextually relevant responses.

  • Prompts

    The practice of designing and optimizing inputs to AI models to achieve desired outputs.