Model Context Protocol (MCP), explained
The protocol docs call it a common way to connect AI apps to other systems. Claude, ChatGPT, and tools such as Cursor can all speak it. You build the connection once. More than one assistant can use it.
On this site, an MCP connector is how an assistant reads a tool you already use, such as a calendar or a customer record. We start with read-only access. A change to a record, a sent message, or spent money waits for a person to say yes.
What does MCP connect?
Data, such as files or a database. Tools, such as search. And set workflows. The assistant asks. The connector does the narrow job you allowed. It is not a pass to every system you own.
Do I need a new connector for every AI product?
That is the problem MCP is meant to shrink. A custom link for each assistant is the old way. A shared protocol means one connector can serve more than one assistant. If your tool already offers a connector, use that and build only what is missing.
Questions
What is MCP in simple terms?
MCP is an open standard that lets an AI assistant plug into your tools through one connector, instead of a custom link for each AI product.
Can the assistant change records on its own?
Only what you allow. We start read-only. Anything that changes a record, sends a message, or spends money asks a person for a yes first.
Which assistants can use it?
The protocol docs name assistants such as Claude and ChatGPT, and tools such as Cursor. We check the one you use before we build.
Where we use this term
Sources
- What is the Model Context Protocol (MCP)?. Model Context Protocol. .
Definition checked against these sources and our own pages on .