Use the MCP Server

Query CAST Imaging results from your own AI client using the standalone CAST Imaging MCP Server

Overview

As well as the AI options built into the CAST Imaging UI, you can interface CAST Imaging with AI clients such as GitHub Copilot and Claude Desktop using the standalone CAST Imaging MCP Server. This is an additional AI option for end-users.

How does it work?

An MCP (Model Context Protocol) server is a standardized way for AI applications to connect to external data sources and tools, acting as a secure bridge between the AI and external systems. The CAST Imaging MCP Server exposes CAST Imaging results to your AI client via the CAST Imaging API. This lets you ask natural-language questions about your application architecture, transactions, quality insights and code objects from within your own AI client, rather than from the CAST Imaging UI.

Who is it for?

This option is best suited to users who already work with an AI client and want to combine CAST Imaging insight with their existing AI-assisted workflow. Configuration is performed outside CAST Imaging and requires the MCP Server to be installed and connected to your AI client.

Getting started

To set up and use the MCP Server, see Getting started with CAST Imaging MCP Server, which covers:

  • Installation on Microsoft Windows.
  • Installation on Linux via Docker.
  • Frequently asked questions about security and data safety.
  • Meta-tool mode.