What it does
The SeatCanvas MCP server exposes three small, read-only tools. They never change your data — they answer a question or compute a layout and return a link. Everything you build still lives in the planner; the assistant is just a faster front door to it.
✦ Get pricing
Returns the current SeatCanvas pricing — the free tier and the one-time Event Pass — straight from the source of truth.
✦ Estimate capacity
Give a room width and height; it returns how many guests fit at your chosen table type, with realistic aisles and spacing.
✦ Create a plan
Give a guest count and event type; it returns a link that opens the canvas already seeded with tables and the venue outline.
Why connect an assistant?
Most people start a seating chart with a vague question: how many tables do I need, will everyone fit, where do I even begin? Answering that in a chat you already have open removes the blank-canvas problem. You describe the event once, and the assistant hands back a concrete starting point — the right number of tables, sized to your room, with the head table in place for a wedding.
Because the layout the assistant promises is computed by the exact same code the canvas renders, what you see in the chat is what you get when the link opens. From there it is the full SeatCanvas planner: drag tables, import your guest list, assign seats, and export. No account is required to look; sign in only when you want the plan saved across your phone and laptop.
Connect your assistant
Point any MCP client at this endpoint. Connector availability changes often — confirm the steps in your client's current docs.
Claude
Settings → Connectors → Add custom connector, then paste the URL:
https://www.seatcanvas.app/api/mcpChatGPT
In Developer Mode, add an app/connector pointing at the URL (availability moves fast — confirm at openai.com):
https://www.seatcanvas.app/api/mcpCursor
Add an entry to your mcp.json:
{ "mcpServers": { "seatcanvas": { "url": "https://www.seatcanvas.app/api/mcp" } } }Also listed on Smithery.
Frequently Asked Questions
What is MCP?
The Model Context Protocol is an open standard that lets AI assistants call external tools. SeatCanvas exposes three read-only tools over MCP so an assistant can answer seating questions and hand you a ready-to-edit plan.
Which assistants can use it?
Any MCP-capable client. We document Claude (custom connectors), ChatGPT (developer-mode apps), and Cursor. Availability of each connector type changes often, so check the client's current docs.
Do I need an account to use the tools?
No. The tools return a link that opens a prefilled plan on the canvas anonymously. You only sign in if you want to save it across devices.
Is my data sent anywhere?
The tools are read-only and deterministic — they build a URL from the numbers you give. We log anonymous, non-personal usage counts to decide whether to keep the integration.