Model Context Protocol
An open standard, usually called MCP, for connecting AI models and agents to external tools and data sources in a consistent way, so they can discover and call an organization's capabilities.
The Model Context Protocol, or MCP, is an open standard that defines how AI applications connect to external tools and data. Instead of writing a custom integration for every pairing of model and system, an organization can expose a capability through an MCP server, and any compatible AI client can discover and use it.
In practice an MCP server describes a set of tools (actions an agent can call), resources (data it can read) and prompts, along with their inputs and outputs. The agent's client then offers those to the model.
Why it matters
Agents do not need your application's screens; they need its capabilities. A standard for exposing those capabilities makes the "callable surface" of a business much easier to build, and makes tool design a product decision. It does not, however, make a capability safe to hand over. Whether an action is ready for an agent depends on its meaning, permissions, side effects and recovery path: this is capability readiness. And because capability is not authority, what an agent may do with an exposed tool still needs explicit rules.
MCP is a building block for agentic AI.
Read more in The next AI interface may never be seen.
Related terms
Capability readiness
How ready an organization's actions are to be used by AI agents: whether each action is a real business capability, exposed cleanly, with clear permissions, side effects and recovery.
Agentic AI
AI systems that pursue a goal by planning, using tools and taking multi-step actions with some autonomy, rather than only answering a single prompt.
Capability is not authority
A design principle for AI agents: being technically able to perform an action does not mean the agent should be permitted to perform it. Delegation needs gradients of authority.