drawio-skill
Use when the user requests diagrams, flowcharts, architecture diagrams, ER diagrams, UML / sequence / class diagrams, SysML / MBSE diagrams (block definition, internal block, requirement, parametric), BPMN business process diagrams, swimlane / cross-functional flowcharts, network topology, cloud architecture from Terraform or Kubernetes manifests, ML/DL model figures (Transformer/CNN/LSTM), mind maps, or any visualization. Also use proactively when explaining systems with 3+ components, complex
Security Assessment
About drawio-skill
The Draw.io Diagrams skill generates .drawio XML diagrams and exports them to PNG, SVG, PDF, and JPG locally using the native draw.io desktop application CLI, with no browser automation required. It solves the problem of producing precise, polished, richly-styled technical diagrams — architecture, network, strict UML, ER diagrams, SysML/MBSE, BPMN, swimlanes, cloud topology, and ML/DL model figures — while keeping the exported images editable (PNG/SVG/PDF exports support embedding the full diagram XML so the file reopens as an editable diagram).
The skill bundles roughly 31 helper scripts and a set of reference documents, grouped by use case: authoring XML, importing code (Python, Go, JavaScript, Rust, Docker, Compose), importing infrastructure-as-code (Terraform, Kubernetes), importing API specs (OpenAPI), rendering live infrastructure, comparing and annotating diagrams, and reverse-exporting. A shape search index of over 10,000 official draw.io shapes (AWS, Azure, GCP, Cisco, Kubernetes, UML, BPMN, P&ID, electrical) lets it resolve exact style strings instead of guessing. It supports learnable style presets, optional Graphviz-based auto-layout, and a vision-model self-check step that is gracefully skipped when unavailable. The live-infrastructure recipes read local tool state from terraform, docker, and kubectl that the user has already authenticated — the skill parses only the JSON those tools print and never reaches out to a cloud account itself.
It targets engineers, architects, and technical writers on macOS, Linux, or Windows who need diagrams-as-artifacts with strict styling and exportable output. It explicitly routes casual whiteboard sketches to excalidraw/tldraw and diagrams-as-code to mermaid/plantuml.
FAQ
What do I need installed to use it?
The draw.io desktop app CLI must be on your PATH (macOS, Linux, or Windows). Optional auto-layout via scripts/autolayout.py additionally needs Graphviz (the dot binary). The self-check step needs a vision-enabled model and is skipped if unavailable.
Can the exported images be edited later?
Yes. PNG, SVG, and PDF exports support --embed-diagram, so the exported file contains the full diagram XML and reopening it in draw.io recovers the editable diagram. Double extensions like name.drawio.png signal embedded XML.
Does it connect to my cloud account to draw live infrastructure?
No. The live-infra recipes only parse JSON printed by local tools you already authenticated (terraform show -json, docker inspect, kubectl get -o json). The skill never reaches out to a cloud account itself; if a tool or its auth is missing it falls back to the declared-config importer.
When should I choose something else?
For a hand-drawn or whiteboard look use excalidraw or tldraw; for diagrams-as-code that live in git and render in Markdown use mermaid or plantuml; for freeform infinite-canvas sketching use tldraw.
All Files
60 filesInstall drawio-skill
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
agents365-ai/drawio-skill