Overview

What is MCP?

MCP (Model Context Protocol) is an open-source standard for connecting AI agents to external tools and data sources. Rather than requiring custom integrations for each tool-LLM pair, MCP provides a universal protocol that allows AI agents to discover and invoke capabilities through a standardized interface.

MCP follows a client-server architecture where an MCP host establishes connections to one or more MCP servers. For a more in-depth view of the MCP architecture, please refer to the official MCP documentation.

How the DVT MCP Server Fits In

The DVT MCP Server exposes tools for accessing project information, enabling AI agents to work with accurate, compiler-backed knowledge, rather than partial or inferred understanding.

Note

The MCP Server can also be used by AI agents in DVT IDE. For more information, please refer to the official DVT IDE documentation for Eclipse or VS Code

How it Works

To set up the DVT MCP Server within an AI agent, an MCP Connector needs to be configured within the agent’s MCP configuration. The MCP Connector establishes connection to a DVT MCP CLI Endpoint, which actually compiles the code.

Upon receiving tool calls, the MCP Connector forwards requests to the DVT MCP CLI Endpoint serving the given project.

The ensemble of these 2 components represents the DVT MCP Server.