Glossary

MCP Server

An MCP server is a service that implements the Model Context Protocol to expose tools, resources, and prompts to AI assistants. It handles connection management, capability negotiation, and request routing between AI clients and your underlying systems.

Explanation

An MCP server sits between your existing services and AI assistants. It translates your API endpoints, database queries, and business logic into MCP primitives that AI models can understand and use. The server handles protocol negotiation (what capabilities are available), transport (HTTP for cloud, STDIO for local), authentication (who can access what), and session management (maintaining state across interactions). Building an MCP server with xmcp involves defining your tools, resources, and prompts as TypeScript functions with type-safe schemas, then deploying the server. The framework handles all protocol-level concerns — you focus on the business logic. A single MCP server can serve multiple AI clients simultaneously, making it an efficient way to AI-enable your entire product surface.

Related Terms

Ready to build your own MCP server?

Take the 2-minute quiz to find out if MCP is the right fit for your project, then get a step-by-step plan to ship it.