What is the Model Context Protocol (MCP)?
Quick answer
The Model Context Protocol, or MCP, is an open standard that lets AI applications connect to outside tools and data in one common way. A tool maker builds an MCP server once, and any AI app that speaks MCP can use it, instead of everyone writing a custom connector.
Last updated
Updated · By Robert Breen
Why it matters for a small business
An AI assistant is only as useful as what it can reach. Before MCP, every pairing of an AI app and a business tool, such as a chat assistant and your file storage, needed its own custom connection. Anthropic introduced MCP as an open protocol in November 2024 to replace that patchwork with one shared way of describing tools, data and how to call them. Other AI companies and software makers have since added support, so the same connector can work in more than one app.
For an office team, MCP mostly shows up as a setting: an assistant or automation tool offers to connect to your calendar, CRM or documents, and behind the scenes it talks to an MCP server. The questions that matter are the usual ones. What can this connection read? What can it change? Who approved it? A connector that can send email or delete files deserves the same care as giving a new hire those permissions.
In a real lesson: n8n AI Agent Tutorial: Save Social Media Ideas to Google Sheets
None of Stepthrough's live lessons use MCP itself, but the n8n AI agent lesson shows the problem it solves. You build an agent for BrightPath Marketing, a made-up marketing agency, and give it one tool: a Google Sheets Tool pointed at a sheet called Social Media Ideas, set to Append Row, with columns Platform, Post Topic, Hook, Description and Call to Action.
The system message says "When the user asks you to save the ideas, use the Google Sheets tool to store each idea." When you type Yes, save those ideas to the sheet, the agent reads the tool's description, decides to call it and writes three rows. That is the pattern MCP standardizes: a tool that describes what it does, and a model that decides when to use it.
Here the tool is built into n8n, so the connection only works inside n8n. With MCP, a Sheets connector written once as an MCP server could be offered to any app that supports the protocol. n8n has added MCP nodes for this; check n8n's docs for the current names and setup.
Try this lesson free or read the step-by-step guide.
Common confusions
MCP vs an API
An API is how one program talks to a specific service. MCP is a common wrapper on top that describes tools in a way AI apps understand, so the model can find and call them. Many MCP servers call an ordinary API underneath.
MCP vs tool calling
Tool calling is the model's ability to request a tool. MCP is a standard way to offer tools to that model from outside the app. One is the skill, the other is the plug.
Tips
- Only connect MCP servers from sources you trust, the same way you would vet a browser extension.
- Start with read-only connections, and add write access once you have watched what the assistant does with it.
- Treat text that comes back through a connector as data, not instructions. Prompt injection can arrive through tools too.
Related terms
More AI agents terms
Where you use it: free lessons
- n8n AI Agent Tutorial: Save Social Media Ideas to Google Sheets (n8n, 12 min)
- AI Email Responder: Draft Gmail Replies Automatically with n8n (n8n, 12 min)
- Lovable + n8n: Build a Web App with an AI Backend (Lovable and n8n, 15 min)
Frequently asked questions
- Who created MCP?
- Anthropic introduced the Model Context Protocol in November 2024 and published it as an open standard, so other companies can build servers and clients for it.
- Do I need to know MCP to build an n8n agent?
- No. n8n's built-in tool nodes, like the Google Sheets Tool, cover most beginner builds. MCP becomes useful when you want to reuse the same tools across several AI apps.