About invitas
invitas is an MCP connector for managing digital invitations and event details through conversational clients. Hosts can work with events, guests, RSVPs, seating, and related communications using its hosted service.
Key features
- 01Create and manage events
- 02Customize invitations and event sections
- 03Manage guest lists and RSVPs
- 04Assign guests to tables
- 05Moderate messages and prepare reminders
Use cases
- 01Organize a wedding guest list
- 02Update event details and invitations
- 03Track RSVPs and seating arrangements
MCP support
The hosted MCP server exposes 31 tools for events, guests, seating, and communications over Streamable HTTP. It supports OAuth 2.1 with PKCE or personal access tokens.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://invitas.co/mcpclaude mcp add --transport http invitas-co-invitas https://invitas.co/mcp{"mcpServers":{"invitas-co-invitas":{"type":"http","url":"https://invitas.co/mcp"}}}Open in Cursor{"servers":{"invitas-co-invitas":{"type":"http","url":"https://invitas.co/mcp"}}}Open in VS Code{"mcpServers":{"invitas-co-invitas":{"type":"http","url":"https://invitas.co/mcp"}}}{"mcpServers":{"invitas-co-invitas":{"serverUrl":"https://invitas.co/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEinvitas MCP connector

Official Model Context Protocol (MCP) server for invitas.co. Create, customize, and manage your invitations, guest rosters, RSVPs, and table seating through conversation with Claude, ChatGPT, Cursor, and any MCP-compliant client.
Connector Endpoint: https://invitas.co/mcp (Streamable HTTP / JSON-RPC 2.0)
Architecture & Interaction Flow
The connector exposes 31 native tools via Streamable HTTP and JSON-RPC 2.0, authenticating requests via OAuth 2.1 (PKCE) or Personal Access Tokens (PAT).
sequenceDiagram
autonumber
actor User as Host / Developer
participant Client as MCP Client (Claude / ChatGPT / Cursor)
participant Connector as invitas MCP Server (https://invitas.co/mcp)
participant App as invitas.co Core API
User->>Client: "How many guests confirmed for my event?"
Client->>Connector: POST /mcp (tools/call: list_guests)
Connector->>App: GET /api/v1/events/{id}/guests (Bearer Token)
App-->>Connector: 200 OK (guests & RSVPs payload)
Connector-->>Client: 200 OK (JSON-RPC tool result)
Client-->>User: "You have 26 confirmed guests out of 41 invited."
Tool Categories (31 Tools)
All tools are automatically catalogued from the core API specification. See Full Tool Catalogue for schemas and parameter details.
| Category | Available Tools | Required Scopes |
|---|---|---|
| Discovery & Guidance | guide, search, fetch | events:read |
| Events & Templates | list_templates, list_events, get_event, create_event, update_event, publish_event, unpublish_event, duplicate_event, export_event | events:read, events:write |
| Media & Sections | add_event_image, upsert_custom_section, set_custom_fields | events:read, events:write |
| Guests & RSVPs | list_guests, get_guest, add_guest, update_guest, remove_guest, import_guests, export_guests | guests:read, guests:write |
| Seating & Tables | get_seating_chart, update_seating_chart, assign_guest_table, unassign_guest_table | seating:read, seating:write |
| Comms & Wall | list_wall_messages, moderate_wall_message, prepare_reminder, get_event_statistics | events:read, guests:read |
Client Configuration Examples
Claude Desktop (claude_desktop_config.json)
{
"mcpServers": {
"invitas": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-sse",
"https://invitas.co/mcp"
]
}
}
}
Cursor (.cursor/mcp.json)
{
"mcpServers": {
"invitas": {
"url": "https://invitas.co/mcp"
}
}
}
Direct HTTP Request Example
# Discover available tools
curl -X POST https://invitas.co/mcp \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_PERSONAL_ACCESS_TOKEN" \
-d '{"jsonrpc": "2.0", "id": 1, "method": "tools/list"}'
Supported Assistants
Step-by-step connection guides for every major client:
- ChatGPT
- Claude app and web
- Claude Code
- Codex CLI
- Gemini CLI
- Cursor
- VS Code (GitHub Copilot)
- OpenCode
- Hermes Agent (Nous Research)
- OpenClaw
- Grok (grok.com)
- Grok Build
- Grok Bot for teams
- Pi coding agent
- Windsurf / Cascade
- Zed
- Muse Code (Meta)
- Other MCP clients and terminal
Sign-in and Permissions
The connector uses Streamable HTTP and OAuth 2.1 with PKCE and dynamic client registration. Sign in to invitas when prompted, then review the consent screen before authorising your assistant.
| Permission | Access |
|---|---|
events:read | Read events. |
events:write | Create and update events. |
guests:read | Read guest information. |
guests:write | Add and update guests. |
seating:read | Read table arrangements. |
seating:write | Update table arrangements. |
Terminal clients that support OAuth can use browser sign-in through the guides above. For clients that support bearer headers instead, create a personal access token in Mi cuenta, choose its permissions and send it in the Authorization: Bearer <token> header. See Other MCP clients and terminal for connection details.
What the Assistant Sees, and How to Disconnect
Within the permissions you authorise, your assistant can access the event information you see in your dashboard, including guest contact details, invitation links, wall messages, tables and prepared reminders. When you connect and authorise access, that information can be shared with your assistant provider, such as Anthropic for Claude or OpenAI for ChatGPT. To end access immediately, open Mi cuenta, select Apps conectadas, then Desconectar; changes already made are kept. See Data and permissions for details.
Examples
See example conversations for requests you can make, the tools involved and the results to expect.
Links
- Assistants
- Developers and REST API · OpenAPI specification
- Privacy policy · Sub-processors · Terms
- Help · Support: [email protected]
Repository Development
This repository contains documentation and a generated tool catalogue; the hosted connector runs in the invitas application. With Node.js 22 installed, run npm run check to verify that generated catalogue files match their source. Run npm run catalog after editing docs/catalog.source.json to regenerate them. See CONTRIBUTING.md.
The docs/ directory contains the catalogue and client guides, examples/ contains conversations, and scripts/ contains the catalogue generator. Changes to this repository publish documentation through the repository; deploying the hosted connector is handled by the application repository.
License
Documentation: CC BY 4.0.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.