What is MCP (Model Context Protocol)?
MCP is the standard that lets your AI assistant reach beyond the chat and actually do things, like publishing a website.
On this page
TL;DR
MCP, the Model Context Protocol, is an open standard that lets AI assistants connect to external tools through connectors. Once you add a connector, your assistant can use its actions from the conversation. SayLive is an MCP connector that publishes websites.
The short definition
The Model Context Protocol, MCP, is an open standard for connecting AI assistants to outside systems. Instead of being stuck with only what it was trained on, an assistant with MCP can call tools, read resources and run actions that a connector exposes.
How a connector works
A connector is a small service that speaks MCP. It tells the assistant which actions it offers, and the assistant calls them when you ask. You add a connector once, authorise it, and it appears alongside the assistant's built-in abilities.
- Tools: actions the assistant can run, such as "publish this site"
- Resources: data the assistant can read, such as your list of sites
- Prompts: ready-made instructions a connector can offer
What you can do with MCP
MCP is general. Connectors exist for files, databases, design tools, project trackers and more. The registry already lists thousands of servers. Publishing a website is one concrete example: the connector takes the files your assistant wrote and puts them online.
With an MCP publishing connector, "make me a site" and "put it online" happen in the same conversation.
MCP and SayLive
SayLive is an MCP connector focused on one job: publishing. You add it to a compatible assistant such as Claude Desktop, and from then on your agent can publish, update and roll back sites by talking to SayLive over MCP.
If you want to see what that looks like in practice, how it works walks through a publish, and the docs cover adding the connector.
Frequently asked
What does MCP stand for?
MCP stands for Model Context Protocol. It is an open standard that lets AI assistants connect to external tools and data through connectors.
What can you do with MCP?
You can give an AI assistant new abilities by adding connectors. Depending on the connector, it can read files, query databases, manage projects or, in the case of SayLive, publish websites, all from the conversation.
Do I need to be technical to use an MCP connector?
No. Adding a connector in an app like Claude Desktop is a few clicks and an authorisation. After that you just ask the assistant to use it in plain language.