MCP: The Standard Plug Into Your Tools

In short

An MCP is a standard plug: it connects Claude Code to the tools you already use.
The signal you need it: you keep finding yourself copying data into the chat from another window.
Ready-made ones already exist for the most widely used tools, and you add them in a few steps.
Only connect what you trust: whatever comes in from outside enters the conversation.

The limit of everything you've seen so far

Up to this point Claude Code lives inside one folder. It reads your files, edits them, runs commands. But your work doesn't all fit in one folder: there's email, the calendar, documents in the cloud, the management software, the place where you track things to do.

As long as those worlds stay outside, it's on you to be the bridge: you open the other window, copy, paste into the chat, read the reply, copy it back over. You're the bridge, and it's a job nobody wants.

The standard plug

The problem of connecting different tools is an old one: each has its own way of being talked to, and every connection would have to be custom-built.

MCP solves it by agreeing on a single format — it's an open protocol, not an Anthropic thing. Like an electrical outlet: if your tool exposes that plug, Claude plugs into it with no custom work.

In practice: whoever builds a service makes their connection available, you add it, and from then on Claude can read and act on that system the same way it does with your files.

What you actually do with it

Ask something that's sitting in the management software without opening it and without copying anything.
Have a reply drafted straight from the emails that came in.
Pull numbers from an archive or a database and put them into the format you use.
Have a job's record updated in the place where you keep track of things.

The way to tell if you need it is always the copy-paste: if you find yourself going back and forth between two windows, that bridge needs to go.

How you add one

You don't have to build anything: for the most widely used tools the connections already exist. There's a directory of verified connectors to start from, and in the desktop app you add them from the connectors screen without touching any configuration file. From the terminal there's a dedicated command.

From there on, the connection is simply one more tool the model can ask for — under the same rule as always: it asks, and the tool only runs if you allow it. The permissions you learned in chapter 5 apply here too.

What's next

Now you have skills, helpers, instructions and connections. That's a lot of stuff built by hand — and the natural question is: how do I carry it to another computer? How do I give it to a colleague? In the next lesson, plugins.

Did you get it?

Two minutes, no grade. Trying to answer — even getting it wrong — cements the concepts far more than rereading. Not in the mood? Skip it: the lesson still counts as complete.

What is an MCP, in one sentence?

What's the signal that you'd need a connection?

Why do you need to be careful about what you connect?

1
1. 1
2. 2
3. 3
2
1. 4
2. 5
3. 6
4. 7
3
1. 8
2. 9
3. 10
4. 11
5. 12
4
1. 13
2. 14
3. 15
4. 16
5. 17
6. 18
5
1. 19
2. 20
3. 21
4. 22
5. 23
6. 24
6
1. 25
2. 26
3. 27
4. 28
5. 29
6. 30
7
1. 31
2. 32
3. 33
4. 34
8
1. 35
2. 36
3. 37
4. 38
9
1. 39
2. 40
3. 41
10
1. 42
2. 43
3. 44
11
1. 45
2. 46