DayuanJiang/next-ai-draw-io

What it solves

Next AI Draw.io simplifies the creation and modification of complex diagrams by replacing manual drag-and-drop drawing with natural language commands. It eliminates the tedious process of manually arranging shapes and connectors, especially for technical architectures, by allowing users to describe what they want to visualize.

How it works

The application is built with Next.js and the Vercel AI SDK to interface with various LLMs. It translates user prompts into the specific XML format required by draw.io. The system can also ingest images, PDFs, and text files to extract content for diagram generation. For advanced users, it provides an MCP (Model Context Protocol) server, allowing AI agents like Claude Desktop or Cursor to generate diagrams directly in the browser.

Who it’s for

It is designed for software architects, developers, and technical writers who need to create cloud architecture diagrams (AWS, GCP, Azure), flowcharts, and system visualizations quickly without manual drawing.

Highlights

• LLM-Driven Editing: Create and refine diagrams using a chat interface.
• Image-to-Diagram: Replicate and enhance existing diagrams from uploaded images.
• Document Ingestion: Generate visualizations directly from uploaded PDF or text files.
• MCP Server Integration: Allows AI agents to create diagrams via the Model Context Protocol.
• Cloud Architecture Support: Specialized libraries for AWS, Azure, and GCP.
• Animated Connectors: Supports dynamic connectors for better flow visualization.
• Version Control: Tracks diagram history to allow restoration of previous versions.