Automation glossary
Short, plain definitions of the terms you meet when buying, building or selling automations with n8n, Make, Zapier and AI tools. Terms are listed alphabetically, and each links to related listings or guides.
Jump to a term
Agent
In AI, a system in which a language model decides which steps or tools to use to reach a goal, often in a loop, instead of following a fixed sequence of steps set in advance.
API
Application programming interface: a defined way for one piece of software to ask another for data or actions, most often over HTTP. Automation tools use APIs to read from and write to the apps they connect.
API key
A secret string that identifies and authorises an application when it calls an API. Treat it like a password: store it in the tool’s credential storage, not in a shared file or a message.
Blueprint
Make’s export format for a scenario: a JSON file describing its modules, settings and connections between them. A blueprint can be imported into a new scenario, and a buyer then connects their own accounts.
Bundle
A single listing that packages several related workflows or files together. On this marketplace, bundles have their own category.
Credential
Stored sign-in details that let an automation connect to an app, such as an API key, an OAuth token or a username and password. n8n calls them credentials, Make calls them connections, and Zapier connects app accounts.
Cron / schedule
A schedule that starts a workflow at set times. Cron is a compact text format, typically five fields for minute, hour, day of month, month and day of week. Most automation tools also offer a point-and-click schedule trigger.
Error handling / retry
What an automation does when a step fails. Error handling can send an alert, route the data elsewhere or stop safely, and a retry runs the failed step again, often after a delay. Retries are safest when the step is idempotent.
HTTP request
A call to a web address using a method such as GET or POST. Automation tools provide an HTTP request step so a workflow can talk to any service that has an API, even without a ready-made integration.
Idempotency
A property of an operation: running it more than once with the same input has the same effect as running it once. It matters for retries and for webhooks that may be delivered twice, because it prevents duplicate records or duplicate charges.
JSON import
Loading a workflow from a JSON file or text into the tool. n8n imports workflows from a file, a URL or pasted JSON, and Make imports a scenario from a blueprint file.
LLM
Large language model: a model trained on large amounts of text that generates text in response to a prompt. LLMs power chat assistants and the AI steps used inside many automations.
MCP
Model Context Protocol: an open standard for connecting AI applications to external tools and data sources, so that an assistant can use them in a consistent way.
MCP server
A program that exposes tools, data or prompt templates to AI applications through the Model Context Protocol. It can run on your own machine or be hosted remotely. Review what a server can access before you connect it.
Node
A single step in an n8n workflow, such as a trigger, an app action or a piece of logic. Make calls the equivalent a module, and Zapier calls it a step.
OAuth
An authorisation standard that lets an app act on your account at another service with limited permissions, without ever seeing your password. You approve access on a consent screen, and the access can be revoked later.
Polling
Checking a service at regular intervals to see whether anything new has happened. Polling triggers can be delayed by up to the interval, whereas a webhook is pushed as soon as the event occurs.
Prompt
The instructions and context given to a language model. The quality of the prompt strongly affects the quality of the output.
RAG
Retrieval-augmented generation: looking up relevant documents or data at question time and giving them to a language model together with the prompt, so the answer is based on that material.
Rate limit
A cap on how many requests an API accepts in a given period. Going over it usually returns an HTTP 429 “Too Many Requests” response. Workflows handle it with delays, batching or retries.
Scenario
Make’s name for an automation: modules connected in a flow, started by a schedule, a webhook or an app event.
Self-hosting
Running software on servers you control rather than using the vendor’s cloud service. n8n can be self-hosted; Make and Zapier are cloud services. You then look after updates, backups and security yourself.
Skill
A folder of instructions, with optional scripts and reference files, that Claude loads when it is relevant to a task. The core is a SKILL.md file with a name and description.
Sub-workflow
A workflow that is called from another workflow so logic can be reused. In n8n a sub-workflow is run from the Execute Workflow node. When you buy a workflow that uses one, make sure it is included.
Template
A ready-made workflow, scenario or Zap that you copy and adapt instead of building from scratch. You connect your own accounts and adjust the settings for your case.
Tool
In AI applications, a function a model can call to fetch data or take an action, such as searching a database or creating a calendar event. MCP servers expose tools.
Trigger
The event that starts an automation: a webhook call, a schedule, a new row in a spreadsheet or a new email. Every workflow, scenario and Zap begins with one.
Webhook
A web address that receives a request when something happens in another system. It lets an app push an event to a workflow immediately, instead of the workflow polling for it.
Workflow
A sequence of connected steps that automates a task. It starts from a trigger and runs the same steps each time it fires.
Zap
Zapier’s name for an automation: a trigger followed by one or more actions.