// glossary
MCP Server plain-English.
Operator-grade definition. Plain words, plus where the term shows up in real work.
What is a MCP Server?
An MCP server is a process that exposes a set of tools, resources, and prompts to LLM hosts — like Claude, Cursor, or an agent runtime — over the Model Context Protocol. It is the standard wire between an AI agent and the systems it is allowed to act on.
Before MCP, every agent integration was a one-off: each model host invented its own way to discover and call external tools. The Model Context Protocol is the standardization layer — discovery, schema, invocation, streaming, cancellation, notifications — so the same MCP server can be wired up to many hosts without rewriting it.
A production MCP server is not just the protocol core. It runs behind a transport (stdio, SSE, HTTP), inside an auth and rate-limit layer, with its tool implementations talking to internal APIs, vendor APIs, or a data store. Observability, an inspector for debugging, and a kill switch are what move it from a demo to infrastructure.
The most common failure mode is shipping an MCP server without any of those: no auth, no observability, no kill switch. The protocol is easy; the production surface is what makes the server safe to point a real agent at.
Related questions
Is MCP the same as a plugin or a function call?
Related, not the same. Plugins were vendor-specific; function calling is a model API feature. MCP is a host-agnostic protocol for connecting models to tools and data, designed for reuse across hosts.
Do I need to write my own MCP server?
Only if your tools, data, or internal systems are not already exposed by an existing server. Otherwise compose existing ones. Hand-rolled MCP servers earn their fee where the surface is proprietary.
What language are MCP servers usually written in?
TypeScript and Python dominate, because that is where the official SDKs live. Any language with the SDK or a hand-rolled implementation works.
Related work
Need someone to actually run this in production? book a call.