I started using Draw.io MCP to generate diagrams from structured input and keep them tied to code and infrastructure. Instead of dragging boxes around by hand…

Thomas Thornton is a software engineer and technology enthusiast known for his expertise in web development and cloud computing. Readers can learn about coding best practices, cloud architecture patterns, and emerging technologies. With blog posts, tutorials, and technical analysis, Thomas Thornton shares  insights and best practices for building modern and scalable software solutions.

Thomas Thornton

Draw.io MCP connects draw.io with AI tools via the Model Context Protocol to generate architecture diagrams from structured text, CSV, or Mermaid input. Instead of manually arranging shapes, engineers can produce a draft .drawio, .png, or .svg file in seconds, commit it to Git, and regenerate it when infrastructure changes. The post covers practical usage patterns including a CI/CD diagram example using a GitHub Copilot skill, tips for keeping automated diagrams readable (3-4 lanes, left-to-right flow, one responsibility per box), and a lightweight team adoption path. The core argument is that diagrams should be version-controlled living assets that evolve with code, not static slide-deck artifacts.

Draw.io MCP for Diagram Generation: Why It’s Worth Using

A Real Example: Generating CI/CD Diagrams

Amazing tool, I have been looking for something like this for a while now!

Interesting! Definitely trying this out

Mermaid is WAYYYYYY better than this. <a href="https://mermaid.js.org/" target="_blank" rel="noopener nofollow">https://mermaid.js.org/</a>