Glossary term

Model Context Protocol

An open protocol for connecting AI applications to external tools and data sources.

How it works

MCP is an open standard introduced by Anthropic for connecting AI applications to tools and data sources. An MCP server exposes capabilities such as reading files or querying a database, and any MCP-compatible app can use them.

It aims to replace one-off custom integrations with a common interface.

How MCP connects things

Step 1AI application (client)
Step 2MCP protocol
Step 3MCP server
Step 4Tool or data source

In code

A tool call in the protocol (simplified)

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {"name": "get_weather", "arguments": {"city": "Paris"}}
}

Why it matters

A shared protocol means a tool built once can work across many assistants and agents.

Related terms

Tool use Function calling Agent