This project provides a YAML-first, offline-first system for authoring and exporting engineering, network, and publication-ready diagrams. It targets engineers and researchers needing repeatable diagram generation and management, offering benefits like version-controlled diagram artifacts and diverse import capabilities.

How It Works

The system employs a base skill (skills/drawio) for core diagramming primitives and an optional academic overlay (skills/drawio-academic-skills) for publication-specific policies and assets. Diagrams are defined via .drawio, .spec.yaml, and .arch.json files for local, repeatable editing. Exports can leverage draw.io Desktop for raster/PDF formats or generate embedded SVGs. An optional @next-ai-drawio/mcp-server enables live browser-based refinement, though the core workflow remains offline-first.

Quick Start & Requirements

• Install: npx skills add bahayonghang/drawio-skills or manual installation.
• Prerequisites: Node.js/npm. draw.io Desktop is recommended for raster/PDF/JPG and embedded SVG exports. The @next-ai-drawio/mcp-server is optional for live refinement.
• Docs: VitePress site (implied by repo structure). Examples are provided within the README.

Highlighted Details

• YAML-First Artifacts: Maintains .drawio, .spec.yaml, and .arch.json in sync for reproducible local editing.
• Desktop-Enhanced Export: Integrates with draw.io Desktop for PNG, PDF, JPG, and embedded .drawio.svg outputs.
• Comprehensive Imports: Generates diagrams from Infrastructure-as-Code (Terraform, Kubernetes, etc.), code projects (Python, JS/TS, Go, Rust), and existing .drawio files.
• Rich Iconography: Includes embedded licensed AI/LLM (Lobe, AI), brand (Redis), and Lucide icons for offline use, supporting SysML and BPMN.
• Network Topology: Supports semantic node types and automatic link labeling from metadata.
• Validation & Postprocessing: Offers pre-export validation (structure, layout, quality) and postprocessing operations like Mermaid conversion, relabeling, and HTML export.

Maintenance & Community

No specific community links (Discord/Slack) or contributor details are provided in the README. Development activity is implied by the inclusion of testing and documentation build scripts.

Licensing & Compatibility

The repository is licensed under the permissive MIT License. The optional @next-ai-drawio/mcp-server uses the Apache-2.0 license. Both are compatible with commercial use and closed-source integration.

Limitations & Caveats

Raster and PDF exports necessitate the installation of draw.io Desktop. Live refinement requires separate setup and configuration of the optional MCP server. Certain advanced postprocessing features are noted as deferred. The academic overlay is designed for offline use and does not rely on the live backend.