Model Context Protocol
Also called MCP
An open standard for plugging AI apps into other tools and data, so one connector can work with many assistants.
Every service an AI might use (a calendar, a database, a project tracker) used to need its own custom integration for each assistant. The Model Context Protocol, or MCP, is a shared standard for that plug. A service builds one MCP server, and any assistant that speaks MCP can use its tools: read the data, search it and, in some setups, change it.
Anthropic introduced MCP in 2024, and Claude, ChatGPT and many developer tools now support it. That's convenient, and it's also a door. An MCP server can reach whatever you give it access to, and the model reads its tool descriptions and data as text, so a careless or malicious server can slip instructions in.
For example
A company builds one MCP server for its ticket system; staff can then ask an assistant "which of my tickets are overdue?" without a separate integration for each AI app.
Put it to use
Where you'll meet model Context Protocol in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.