MCP in three primitives: tools, resources, prompts

The Model Context Protocol exists because people got tired of copying and pasting context between applications. Before MCP, connecting a model to your files, tickets or databases meant manual shuttling or a bespoke integration for every pairing of model and system. That does not scale, and it is a governance headache.

MCP's answer is three primitives. Tools are actions a model can take. Resources are data it can read for context. Prompts are reusable, user-defined inputs. I like this design because it is small enough to explain to a non-specialist in a minute, which is rare for an integration standard and a good sign it will last.

The bigger story is that it was open source from the start, which let a large community form quickly. That shifted effort away from proprietary connectors and toward model intelligence and workflow, where the real differentiation sits. For an enterprise architect, a common protocol also means one place to apply authentication, logging and policy instead of dozens.

Small, open standards tend to win; the MCP specification is a good example of why.