When large language models first arrived, the only way to converse with them about a codebase was to copy/paste chunks of code into a text interface.

Model Context Protocol (MCP) Quickstart

submited by

Style Pass

2024-11-25 22:30:03

When large language models first arrived, the only way to converse with them about a codebase was to copy/paste chunks of code into a text interface. We quickly realized that this not enough, and custom integrations emerged to help with loading the additional context. However, these efforts were fragmented and required everyone to build their own integration. This is where the Model Context Protocol (MCP) comes in.

Anthropic introduced the Model Context Protocol (MCP) to replace fragmented, custom-built integrations with a universal protocol, enabling AI systems to interact with both local and remote resources efficiently.

Each MCP server runs as a separate process, and communicates with clients via a JSON-RPC. Example, you can launch the Brave Search server via the following command:

The host is responsible for discovering the capabilities of the servers, and planning how to utilize them to solve end user problems.