How does the skill prevent broken or unreadable diagrams?

It implements a strict validation checklist that ensures mandatory structural cells (ID 0 and 1) are present, IDs are unique, and style properties use exact camelCase naming conventions.

What format does the AI use to generate diagrams?

The skill uses a simplified, uncompressed mxGraphModel XML format. This reduces complexity and ensures the diagram is human-readable and easily imported by draw.io without errors.

Does it support specific shapes like AWS or cloud icons?

Yes, by providing detailed style properties in the prompts, the AI can generate diagrams that utilize specific icon sets and shape libraries available in draw.io.

Can I use this with my own LLM API key?

Yes, the skill supports configuration for Anthropic Claude, Google Gemini, and OpenAI GPT keys, as well as custom LLM endpoints compatible with OpenAI's API.

AI Diagram Generator for draw.io

Name: AI Diagram Generator for draw.io
Author: markus41

bymarkus41

•

Herramientas de Diseño

Generates and optimizes professional draw.io diagrams using AI-powered XML creation and structural validation.

This skill empowers developers and architects to transform natural language prompts into precise, editable draw.io diagrams. By leveraging specialized LLM instructions and strict XML formatting rules, it ensures generated diagrams are syntactically correct, visually organized, and fully compatible with the diagrams.net ecosystem. It handles everything from complex cloud architectures and UML sequence diagrams to ER models, providing built-in validation to prevent common rendering errors and ensuring professional-grade visual documentation.

Características Principales

0111 GitHub stars

02Multi-model support for Claude, Gemini, and GPT via draw.io integration

03Specialized prompt templates for architecture, UML, and entity-relationship diagrams

04Automated XML generation using simplified, uncompressed mxGraphModel formats

05Built-in validation for structural integrity and unique cell ID management

06Precise perimeter and style property mapping for non-rectangular shapes

Casos de Uso

01Refactoring existing diagrams using AI-assisted suggestions and automated structural fixes

02Generating technical documentation and sequence diagrams from code logic or requirements

03Rapidly prototyping cloud architecture and system design visuals from text descriptions

Características Principales

0111 GitHub stars

02Multi-model support for Claude, Gemini, and GPT via draw.io integration

03Specialized prompt templates for architecture, UML, and entity-relationship diagrams

04Automated XML generation using simplified, uncompressed mxGraphModel formats

05Built-in validation for structural integrity and unique cell ID management

06Precise perimeter and style property mapping for non-rectangular shapes

Casos de Uso

01Refactoring existing diagrams using AI-assisted suggestions and automated structural fixes

02Generating technical documentation and sequence diagrams from code logic or requirements

03Rapidly prototyping cloud architecture and system design visuals from text descriptions