clenttGlossary
Glossary

Model Context Protocol

Also known as: MCP

Definition
Model Context Protocol (MCP) is an open standard for connecting AI models to external tools and data sources, letting an LLM discover and call functions provided by separate servers in a structured, secure way.

Model Context Protocol (MCP) was introduced as an open specification for the integration layer between AI models and the tools they need to act on real systems. Before MCP, every product wired its own bespoke tool-calling layer, which made it hard for AI features to reach across products or for new tools to plug into existing AI experiences.

An MCP server exposes a set of tools (functions with schemas), resources (read-only data the model can pull in), and prompts (templated instructions). An MCP client (typically embedded in an AI chat or agent) connects to one or more servers, lets the model discover what tools exist, and routes tool-calling requests to the right server.

For GTM software, MCP matters because it lets a single AI chat call into many specialized systems through a uniform interface. Instead of building a custom integration for every tool the AI needs to touch, the product exposes its capabilities as MCP servers and the AI orchestrates across them.

How Model Context Protocol relates to Clentt

Clentt's AI chat is built on MCP. Plugins (workflows, leads, instantly, and core records) register as MCP servers, expose their actions as tools, and the chat can call them with up to six rounds of tool-calling per conversation.

Related terms

Continue