Back to browse
GitHub Repository

Turn your architecture into a living, queryable knowledge graph - and render it as beautiful auto-laid-out Excalidraw diagrams. An MCP server for Cursor, Claude Code & Windsurf. Offline, no API keys.

125 starsPython

Excalidraw Architect MCP for AI Based IDEs

by Venky1729·Mar 7, 2026·2 points·0 comments

AI Analysis

●●●BangerSolve My ProblemWizardryShip It

LLM-generated Excalidraw diagrams that actually lay out correctly using Sugiyama algorithm.

Strengths
  • Sugiyama hierarchical layout eliminates manual coordinate fixing — solves the core LLM diagram hallucination problem elegantly.
  • Architecture-aware node styling (Kafka gets stream icons, etc.) reduces cognitive load and improves diagram readability immediately.
  • Stateful editing via natural language means you can iterate on diagrams without exporting/reimporting or manual rewiring.
Weaknesses
  • Niche audience: only useful if you're already using Cursor/Windsurf + regularly generating architecture diagrams.
  • No Windows support mentioned; uv dependency for non-pip install may friction some users.
Target Audience

Engineers using AI IDEs (Cursor, Windsurf) who need to generate architecture diagrams programmatically.

Similar To

Mermaid diagram generation in Claude/ChatGPT · Structurizr DSL · ASCII architecture tools (like asciify)

Post Description

Hey Folks, I've been building an MCP server that generates beautiful Excalidraw architecture diagrams fully auto-laid-out, no manual positioning. It's open source, works with any AI IDE (Cursor, Windsurf, etc.).I wanted to share it early because the core layout engine is solid and already producing clean diagrams.

What Problem Am I Solving? AI IDEs generate architecture diagrams as Mermaid or ASCII art. When they attempt Excalidraw, they hallucinate coordinates and boxes overlap, arrows cross through nodes, and you spend more time fixing the diagram than drawing it yourself. LLMs understand what a system looks like, but they have zero intuition for where things go on a 2D canvas.

How Does This MCP Solve It? You describe the components and connections. The MCP runs a Sugiyama hierarchical layout algorithm to compute positions deterministically -- the AI never touches coordinates. It auto-styles 50+ technologies (say "Kafka" and get a stream-styled node), stretches hub nodes like API Gateways to span their services, and routes arrows around obstacles. The result is a clean .excalidraw file in seconds, with stateful editing so you can say "add a Redis cache" without starting over.

Why Not the Official excalidraw-mcp? The official MCP gives the AI a raw canvas and lets it place elements freely -- great for general sketching, but architecture diagrams still end up with overlapping boxes. excalidraw-architect-mcp takes a structured graph as input and runs a proper layout algorithm, so every diagram is overlap-free by design. It also runs fully offline in your IDE with no API keys, outputs version-controllable .excalidraw files, and includes architecture-aware styling that the general-purpose tool doesn't have.

What's the Future? Diagram-from-code (point at a codebase, get an architecture diagram automatically), live sync so diagrams update as code evolves, richer layouts (radial, swimlane, grid), and an expanded component library covering cloud-provider icons and CI/CD tools. The goal: architecture diagrams as a byproduct of building software, not a separate task.

Similar Projects

Developer Tools●●Solid

Layerd – Layered system architecture diagrams with 3D visualization

The layer-first canvas and a clear components palette (Services, Databases, Queues, Line Keys) make expressing front-end/back-end/data separation quick and visual. The 2D↔3D toggle is a neat flourish for showing stack depth and ownership, but without visible integrations (Terraform/Cloud import, export/collaboration, or smart layout) it reads as a nicer-looking competitor to existing diagram tools rather than a workflow disruptor.

SlickNiche GemEye Candy
hexploits
214mo ago
Developer Tools●●Solid

Claude Architect

C4 model diagrams with comment-to-Claude feedback loop through Channels.

Big BrainNiche Gem
hennessywill
201mo ago