← Back to the A-Z glossary
AI glossary · P1

Skill & Agent

A skill is a reusable method. An agent is a system that pushes a mission forward with tools, context, and decisions.

Agent 4 min read Updated 2026-05-22
— Definition

Skill & Agent, in plain words

A skill is a reusable method. An agent is a system that pushes a mission forward with tools, context, and decisions.

Set up the distinction that keeps you from calling everything an agent.

A concrete example

A watch agent can use an editorial summary skill to produce the final recap.

Why it matters

Confusing skill and agent wastes time when you actually want to automate a real routine.

You'll see skills in Claude Code and agents in Codex, n8n, Zapier, or automated watch tools.

Don't mix it up with

Agentic workflow: An agentic workflow is a sequence of steps where the AI plans, uses tools, checks its work, and keeps going until it hits the result.

Orchestration: Orchestration organizes the steps, tools, or agents needed to reach a result.

Common mistakes

  • Calling a simple style instruction an agent.
  • Building an agent when a skill would do.
  • Letting an agent run without stopping criteria.

Quick checklist

  • First I check whether the word names a concept, a tool, a risk, or a metric.
  • I tie it to a concrete case: A watch agent can use an editorial summary skill to produce the final recap.
  • I keep the main trap in mind: Calling a simple style instruction an agent.

Quick questions

What is Skill & Agent in AI?

A skill is a reusable method. An agent is a system that pushes a mission forward with tools, context, and decisions.

Where will I run into Skill & Agent?

You'll see skills in Claude Code and agents in Codex, n8n, Zapier, or automated watch tools.

Which word should I read next?

Start with Agentic workflow, Orchestration, Multi-agent.

Want to keep going in order?

Head back to the full glossary, search a word, then open only the pages that deserve more than a short definition.

Open the AI glossary