Happy slice of mcp pizzaMCP.pizza

drawio-scientific-illustrator

MCP.Pizza Chef: icebird1998

This tool lets your Codex-connected AI assistant open your draw.io desktop app and build a scientific illustration piece by piece, right on your screen, so you can watch every shape, label, and arrow appear as it goes. Give it a reference image — a photo, a PDF page, or a sketch — and it recreates the figure with editable text, arrows, and panels you can still adjust afterward. It requires draw.io desktop, Git, and Node.js installed on your computer, and works specifically with the Codex app — no account or special access code needed.

Files/PDF
Images/Design

Use This MCP server To

Recreate a reference science figure step by step in draw.io Turn a photo of a diagram into an editable drawing Build a multi-panel science illustration with labeled arrows and legends Watch each shape and label appear live as it draws Export my finished figure as a PNG, SVG, PDF, or JPG Redraw a textbook figure with fully editable labels and shapes

README

Draw.io Scientific Illustrator

Important

项目已迁移 / Project migrated

本仓库目前继续保持公开访问，暂不归档或删除，但已停止功能与兼容性更新。后续开发、版本发布和问题修复均已迁移至 scientific-illustrator。建议所有新老用户下载并使用新项目。

推荐下载新项目 / Download the new project: 项目主页 / Project home · 最新版本下载 / Latest release

This repository remains publicly accessible and is not being archived or deleted at this time, but it no longer receives feature or compatibility updates. All future development, releases, and fixes have moved to scientific-illustrator. New and existing users are encouraged to download and use the new project.

中文说明 · English guide · MIT License

A Codex plugin that lets an AI agent draw scientific figures live inside the visible draw.io desktop canvas. You can watch shapes, labels, arrows, styling, and layout appear step by step. The live workflow calls draw.io's own graph API through a localhost-only MCP server; it does not automate the operating-system mouse or keyboard and does not create XML first and merely open it afterward.

Legacy status: this repository remains online for existing users but is no longer updated. Windows was the tested platform for this version. For current development and cross-platform support, use scientific-illustrator.

English guide

What this repository contains

  • drawio-live MCP server: launches/connects to the visible draw.io desktop editor and edits its active graph model in real time.
  • drawio-file-utils MCP server: validates saved .drawio documents and exports PNG, SVG, PDF, or JPG deliverables.
  • A Codex skill: teaches the agent to inspect a reference, decompose it into editable primitives, draw with pacing, visually review sections, refine the live graph, and save only after the visible drawing exists.
  • A repository-local Codex marketplace so the complete plugin can be installed as one unit.

This is therefore both an MCP implementation and a Codex plugin. MCP provides the callable tools; the plugin is the installable package containing the MCP servers, skill, and presentation metadata.

Requirements

  1. Codex desktop app or Codex CLI with plugin support.
  2. draw.io desktop installed locally.
  3. Git.
  4. Node.js available as node when running the MCP servers outside the bundled Codex runtime. Node.js 22 or newer is recommended.

drawio-scientific-illustrator FAQ

Can I use this to recreate a scientific figure from a photo or PDF page?
Yes — attach a PNG, JPEG, SVG, or rendered PDF page as your reference, and it rebuilds the figure step by step inside draw.io, keeping every label, arrow, and panel editable afterward.
Which app does this work in?
It works specifically inside the Codex desktop app or Codex CLI, not Claude Desktop, ChatGPT, or other assistants, and it also requires draw.io desktop installed on your computer.
Do I need to sign up for an account or get a special access code to use this?
No. It runs entirely on your own computer without needing an account or special code, though you do need Codex and draw.io desktop installed first.
How hard is it to set up?
Setup takes some technical comfort — you'll need Git and Node.js installed, and you install the tool by running a short script or a few commands in a terminal, so it suits someone comfortable following technical instructions.
Can I use this to export my finished diagram as an image or PDF?
Yes — once your figure is complete, it can save the draw.io file and export it as a PNG, SVG, PDF, or JPG.
Is this project still being actively updated?
No. The developer has moved active development to a newer project called scientific-illustrator, so if you're starting fresh, that newer version is the recommended pick; this version still works but no longer receives updates.
Does it take control of my mouse and keyboard while it draws?
No. It talks directly to draw.io's own drawing engine on your computer, so your mouse and keyboard stay free for you to use normally at the same time.
What computer does this run on?
This version was tested on Windows; the newer scientific-illustrator project mentioned above offers broader support across Windows, macOS, and Linux.