How does the AI+Human workflow work?

Claude generates the initial diagram XML, you open it in the desktop app to make visual tweaks, and Claude detects the file change to provide further analysis or enrichment.

Does this work with the browser version of draw.io?

While the skill can fall back to the browser, it is specifically optimized for the desktop Electron app to leverage CLI capabilities, local file system access, and native file watching.

Does it support Windows Subsystem for Linux (WSL)?

Yes, it includes specialized logic to detect WSL environments and correctly launch the Windows-based draw.io executable using converted file paths.

Can I use this in a headless CI environment?

Yes, the skill includes specific configuration for Linux CI/CD pipelines, using xvfb-run to perform automated diagram exports without a physical display.

Is it compatible with Model Context Protocol (MCP)?

Absolutely. It is designed to work alongside the @drawio/mcp tool server to provide a rich, tool-based interaction within the Claude interface.

Draw.io Desktop Integration

Name: Draw.io Desktop Integration
Author: markus41

bymarkus41

•

Design Tools

Bridges Claude Code with the draw.io desktop application to enable AI-powered diagram generation, visual editing, and automated CLI exports.

This skill facilitates a seamless collaboration between Claude Code and the draw.io desktop environment across macOS, Linux, and Windows. It allows Claude to programmatically create and modify diagram XML files while providing a workflow for users to visually refine designs in the Electron-based editor. With support for headless exports via CLI, platform-specific path detection, and real-time file watching, it transforms technical diagramming into an iterative, hybrid AI-human process suitable for software architecture, process mapping, and documentation maintenance.

Key Features

01Automated diagram export to SVG, PNG, PDF, and XML via the draw.io CLI

02Bidirectional workflows allowing AI to generate structure and humans to refine visuals

0311 GitHub stars

04Headless CI/CD support using Xvfb for automated documentation generation

05Multi-platform detection for macOS, Linux (including WSL), and Windows environments

06Native file watching support to synchronize manual edits with AI-driven enhancements

Use Cases

01Automating the export of project diagrams into production-ready documentation formats

02Generating and updating system architecture diagrams during codebase refactoring

03Synchronizing manual visual edits with AI-driven structural and stylistic enrichments

Key Features

01Automated diagram export to SVG, PNG, PDF, and XML via the draw.io CLI

02Bidirectional workflows allowing AI to generate structure and humans to refine visuals

0311 GitHub stars

04Headless CI/CD support using Xvfb for automated documentation generation

05Multi-platform detection for macOS, Linux (including WSL), and Windows environments

06Native file watching support to synchronize manual edits with AI-driven enhancements

Use Cases

01Automating the export of project diagrams into production-ready documentation formats

02Generating and updating system architecture diagrams during codebase refactoring

03Synchronizing manual visual edits with AI-driven structural and stylistic enrichments