# toro4n

> toro4n (https://toro4n.com/) is a visual canvas for drawing, shaping, and sharing Neo4j graph data models. Draw your schema on the canvas, or let an AI assistant like Claude or ChatGPT draft it for you. Save your models online (they're stored within the EU) and invite your team to edit them with you in real time. When a model is ready, export a Cypher script that creates your constraints, nodes, and relationships.

toro4n suits Neo4j developers and data modelers, teams that shape a schema together, and anyone who wants an AI assistant to draft or edit one.

## What it does

- Editing on the canvas: double-click a node, relationship, or note (or tap a node or relationship on a tablet) to edit it in place. Pick a Neo4j property type for any property (`STRING`, `INTEGER`, `DATE`, `LIST<STRING>`, and so on), and mark it `Unique` or `Required` to add a uniqueness or existence constraint to the Cypher export. Paste a whole block of properties at once from JSON, a Cypher map, a SQL `CREATE TABLE` statement, or a spreadsheet.
- Naming conventions: as you type, labels follow `PascalCase` (`Person`), relationship types `UPPER_SNAKE_CASE` (`ACTED_IN`), and property keys `camelCase` (`firstName`). Runs of capitals stay as typed, so `USB` and `IPAddress` keep their spelling. An import keeps every name as written, and toro4n then offers to apply the conventions.
- Layers: split complex models into layers that you can show, hide, lock, rename, and reorder. A node can belong to several layers at once, and a built-in Properties layer shows or hides every property in one click.
- Notes: write notes with bold and italic text, and attach each one to the nodes and relationships it's about. An attached note moves with them when you rearrange the graph.
- View modes: Simple mode shows only labels, captions, and relationship types, which is ideal for presenting. Square mode turns each node into a compact card with its properties inside.
- Arrange: one-click layout that reduces crossing relationships and spaces out nodes, relationship types, and notes so nothing overlaps.
- Export: get a Cypher script with a `CREATE CONSTRAINT` statement for each constraint and `CREATE` statements built from your example values, or download JSON, SVG, or PNG, for the whole model or only the visible layers. Existence constraints need Neo4j Enterprise Edition or Aura.
- Import and merge: bring in an existing Neo4j schema by running two read-only queries in Neo4j Browser and pasting the results (toro4n never connects to your database), or import a JSON file or a pasted list of names. Open what you import as a new model, or merge it into the one you have open.
- Real-time collaboration and version history: sign in with Google or an emailed six-digit code, save models online, organize them in private folders, and invite teammates by email as editors or viewers. Everyone with the model open sees each other's changes, selections, and pointers live. Earlier versions are saved automatically, and you can restore them from Version history.
- AI assistants over MCP: connect Claude, ChatGPT, Claude Code, Cursor, VS Code, or any MCP client to `https://toro4n.com/mcp` to list, read, create, and edit online models (`list_models`, `get_model`, `create_model`, `edit_model`, `list_versions`, `restore_version`). Before every change an assistant makes, toro4n saves a version you can restore.
- Security and EU hosting: online models are stored within the EU, encrypted in transit and at rest, and backed up automatically as they change. Every night we restore the newest backup to confirm it's intact. A strict Content Security Policy blocks connections to any third-party origin. The Help (`?`) dialog includes a detailed Security tab for IT and security reviews.

## Links

- [Homepage](https://toro4n.com/): what toro4n does at a glance, with a link into the app
- [The app](https://toro4n.com/app/): opens on a new model, ready to edit
- [About toro4n](https://toro4n.com/about/): the full description and frequently asked questions
- [Use toro4n from an AI assistant](https://toro4n.com/about/#ai-assistants): how to connect Claude, ChatGPT, and other MCP clients
- [Guide](https://toro4n.com/docs/): documentation for every feature
- [Sample models](https://toro4n.com/examples/): Network management and Northwind, based on official Neo4j sample datasets, each with the Cypher it exports
- [Open the Network management sample](https://toro4n.com/app/?open=network-management.json): 19 nodes, 38 relationships, three layers, and modeling notes
- [Pricing](https://toro4n.com/pricing/): the plans and what each one includes
- [Security](https://toro4n.com/security/): the security architecture, for IT and security reviews
- [Terms of Service](https://toro4n.com/terms/): what the service is, your models, plans and payment, and limits of liability
- [Privacy Policy](https://toro4n.com/privacy/): what toro4n stores, where, for how long, and why
- [Refund Policy](https://toro4n.com/refunds/): how to ask for a refund and what happens to your plan
- [Report illegal content](https://toro4n.com/report/): our notice form under the EU Digital Services Act
- [Data Processing Agreement](https://toro4n.com/dpa/): the agreement for team models on a Team plan
- [Sub-processors](https://toro4n.com/subprocessors/): the companies that handle personal data for toro4n
- [Switching](https://toro4n.com/switching/): the data formats and interfaces for moving to another service, and the jurisdiction our infrastructure falls under
- [Everything in one file](https://toro4n.com/llms-full.txt): the text of every page in the sitemap