How to Connect ChatGPT to Your Wedding CRM: Wedy's MCP Explained

A florist finishing a Saturday install doesn't have time to open a laptop and dig through a CRM before answering a bride's texted question about her invoice. The discerning vendors building six and seven-figure businesses already keep ChatGPT or Claude open in another tab: drafting a client email, refining a contract clause, sharpening a caption. But talking to an AI assistant and running a business through one are two very different things. For most CRMs, the AI can only read what gets pasted in. It cannot see real leads, pull real overdue invoices, or send a follow-up email on a vendor's behalf.

That changed with a technical standard called MCP (Model Context Protocol), and it is why a growing number of wedding professionals are asking how to connect ChatGPT to their wedding CRM directly, not settle for a smarter notepad. Wedy Pro built an official, first-party MCP server so vendors on the Elite plan can ask Claude, ChatGPT, Cursor, or Gemini CLI to check their pipeline, draft and send a client email, or assemble a full, branded proposal, all from the chat window already open on their desk.

This guide breaks down what MCP actually is, why every major AI company adopted the same protocol rather than building rival ones, exactly how to connect ChatGPT or Claude to Wedy Pro, and what it looks like to run a refined, modern wedding business from an AI assistant in practice.

What Is MCP (Model Context Protocol), and Why Every AI Company Adopted It

Anthropic open sourced the Model Context Protocol on November 25, 2024, as a standardized way for AI assistants to connect to the systems where a business's real data and tools already live, instead of requiring a one-off custom integration for every app (Anthropic, November 2024). Before MCP, an AI assistant that wanted to check a CRM, a calendar, and a payment system needed three separate, hand-built connections. MCP gave every AI vendor a shared, standard way to plug into the same tools.

What makes MCP notable is not that Anthropic built it. It is that Anthropic's direct competitors adopted it almost immediately. OpenAI added MCP support across ChatGPT, its Agents SDK, and Codex starting in March 2025, with Sam Altman publicly stating that “people love MCP and we are excited to add support across our products” (ChatForest, 2025). Google DeepMind and Microsoft followed, with Microsoft Copilot Studio's MCP support reaching general availability in 2025 (Microsoft, 2025). By December 9, 2025, Anthropic donated MCP's governance entirely to a new, vendor-neutral Agentic AI Foundation under the Linux Foundation, with OpenAI, Google, Microsoft, AWS, Cloudflare, and Bloomberg signing on as founding members (Linux Foundation, December 2025).

Direct competitors choosing to co-govern the same open protocol, rather than build rival standards, is the strongest signal that MCP is now permanent infrastructure for how AI connects to business software, not a feature that belongs to one company. The adoption curve backs that up: MCP grew from roughly 100,000 downloads in its first month to more than 97 million monthly SDK downloads and over 10,000 active servers within about a year of launch, with first-class support in Claude, ChatGPT, Cursor, Gemini, Microsoft Copilot, and VS Code (AI2Work, December 2025).

The Real Distinction: An AI That Answers Questions vs. an AI That Runs Your Business

Here is the part that matters for a wedding vendor deciding whether this is worth setting up: for most of 2025, MCP connections were read-only. An AI assistant could look something up but could not act on it. That changed on September 10, 2025, when OpenAI rolled out full read and write MCP support in ChatGPT's Developer Mode, meaning ChatGPT could finally write actions back into a connected system, not just fetch information from it (InfoQ, October 2025).

That is the difference between asking an AI assistant a question and letting it operate your business. A read-only connection can tell you how many leads you have. A read and write connection can create the lead, add a note to it, draft a follow-up email, and (with your confirmation) send it. Wedy Pro's MCP server, live at mcp.wedypro.ai, is built for that second category. It exposes more than 43 tools covering projects, leads, Smart Documents, calendar, payments, email, and tasks, so an Elite-plan vendor can genuinely manage day-to-day operations from an AI chat interface instead of logging into the CRM at all.

In practice, that looks like typing “show me all my overdue invoices” into Claude and getting back a real, live breakdown grouped by aging bucket (1 to 30 days, 31 to 60 days, 60-plus), pulled directly from your Wedy Pro account, not a guess based on a screenshot you pasted in.

How to Connect ChatGPT, Claude, or Cursor to Wedy Pro

Connecting an AI assistant to Wedy Pro takes a few minutes and does not require any technical setup, because AI clients like Claude and ChatGPT auto-register themselves using a standard called Dynamic Client Registration. There is no API key to generate or developer console to configure.

1. Open Claude, ChatGPT, Cursor, VS Code, or Gemini CLI and ask it a Wedy-related question or request an action (for example, “connect to my Wedy Pro account”).
2. The AI assistant detects Wedy's MCP server and opens a secure login window at wedypro.ai's authorization page.
3. Log into your Wedy Pro account if you are not already signed in. If you manage more than one vendor account, select the one you want to connect.
4. Review the consent screen, which clearly states what the AI assistant is requesting access to, then select “Allow.”
5. You will see a “Connected successfully” confirmation, and you can close the tab. The AI assistant now has a secure, revocable connection to your account.
6. Start asking it to check your pipeline, review overdue payments, draft a proposal, or check your calendar. The connection persists, so you will not need to reauthorize every session.

Wedy Pro's MCP server is available on the Elite plan. Vendors on the Pro plan or the free tier will see a clear, readable message directing them to upgrade if they attempt an AI tool call, rather than a confusing error.

What You Can Actually Do From the Chat Window

The point of connecting ChatGPT or Claude to a wedding CRM is not novelty. It is removing the friction that keeps a photographer editing galleries at midnight instead of following up on a Sunday inquiry, or a caterer between tastings from ever opening a laptop at all. A few concrete examples of what Wedy Pro's MCP server supports today:

• Check your pipeline and revenue. Ask for your projects by stage, your overdue payments with aging buckets, your completed payments, or a revenue chart for a specific date range.
• Follow up with leads intelligently. Ask the AI to pull up a lead's email thread and draft a follow-up. Before anything is sent, the AI asks for your explicit confirmation. Sending and replying are the only actions marked as requiring that extra step, because they are irreversible.
• Build a full Smart Document. Ask it to create a proposal for a specific client, and it can generate the cover page, contract, and invoice, populate line items and payment schedule, and pull in your saved branding, fonts, and images, all in the same conversation.
• Check your calendar and scheduler availability. Ask what is on your calendar this week, or whether a specific date is open, and it pulls from your synced calendar and Wedy Pro project dates together.
• Manage tasks and notes. Create a task, add a note to a project, or check what is overdue on your task board without opening the CRM.

Picture a wedding photographer wrapping a Saturday shoot who taps out one line to Claude from her phone: “draft a follow-up to the Martinez inquiry and let me approve it before it sends.” By the time she is back at her car, a personalized reply is waiting for her approval, written around the couple's actual questions, not a generic template a DJ or planner would send to every lead.

This is a meaningfully different category from the workflow automation most CRMs offer. A rule-based automation runs a fixed script: lead submits a form, send template A. Wedy Pro's AI agents read the actual intent behind a request, decide what needs to happen, and execute it, whether that is drafting a personalized response, assembling a multi-page proposal, or surfacing exactly the overdue invoices that need attention today. It is the most advanced automation available in a wedding CRM, because it thinks through the request instead of following a fixed script.

Why No Other Wedding CRM Has This (Yet)

As of mid-2026, neither HoneyBook nor Dubsado has an official MCP server or a first-party ChatGPT or Claude integration. The only AI-assistant connections available for either platform are unofficial, third-party connectors built by outside developers, not the vendors themselves. One publicly listed HoneyBook connector explicitly discloses that it is “not affiliated with, endorsed by, or in partnership with HoneyBook, Inc.,” and notes that HoneyBook's own Terms of Service prohibit the kind of data access these unofficial tools rely on (crossaitools.com, 2026). Dubsado's only public AI-facing asset is a static reference page built to help AI systems cite accurate product facts. By its own description, it is not an integration mechanism at all (Dubsado, 2026). 17hats, the third major competitor in this category, is explicitly listed in 2026 industry roundups as having no real AI features (usecarly.com, 2026).

That gap matters because of where the industry is heading. Independent CRM comparison research already ranks native, vendor-built MCP servers above third-party middleware connectors, which rank above no integration at all, citing MCP's 97 million SDK installs as evidence it has become the standard way businesses connect AI to their software (Clarify, 2026). Wedy Pro built and owns its MCP server. Vendors on HoneyBook or Dubsado who want this today have to route around their own CRM's terms of service to get a version of it from an unofficial third party.

Is It Safe to Connect an AI Assistant to Your Business Data?

This is a fair question, and it deserves a direct answer rather than a dismissal. Security researchers have identified real risks with MCP connections generally: over-scoped access that exposes more data than intended, “tool poisoning” where malicious instructions are hidden inside a tool's own description, and lingering access from AI connections that are never revoked (Checkmarx, 2026). Those risks are real for self-hosted or loosely built MCP servers. They are exactly why Wedy Pro built its MCP server as a vendor-controlled, plan-gated product rather than an open connector.

Wedy Pro's MCP server runs on OAuth 2.1 with PKCE, the current security standard for authorization. Every tool call is authorized with a short-lived token that expires in 30 seconds and is never stored or reused. Each session is cryptographically bound to your specific account, so cross-account access is not possible even if a session ID were somehow exposed. Email tools that read your inbox strip HTML into plain text and mark external content clearly, guarding against hidden instructions. Sending or replying to an email always requires your explicit confirmation first. Smart Documents can only be edited by an AI assistant while still in draft, a safeguard that prevents any AI action from overwriting a signature or a payment that already exists on a sent document. And the entire feature is gated to the Elite plan, so it is never an accidental, unscoped connection anyone can stumble into.

Why Wedy Pro Is the Clear Choice for AI-Native Wedding Business Management

Wedy Pro is a full, standalone CRM built for vendors who run polished, elevated client experiences without losing an evening to admin. Everything HoneyBook and Dubsado offer, lead forms, contracts, invoices, payments, automations, scheduling, Wedy Pro offers too, sent from a vendor's own connected email address, never a generic platform address. The MCP server is not a replacement for that CRM. It is an additional, refined way to operate it, one that puts an entire business inside the AI chat window already open every day.

What separates Wedy Pro is that this was built as a first-party product feature, not bolted on. The AI agents that power it are the same agents already reading lead intent and selecting the right email template automatically, the same agents building a client's proposal with the right design blocks pulled in. MCP simply extends that intelligence to whichever AI assistant a vendor already prefers, whether that is Claude, ChatGPT, Cursor, or Gemini CLI.

It is also worth understanding why Wedy is positioned to move first here. Wedy, featured on Shark Tank and in Forbes, and backed by J.P. Morgan, scaled nationwide on the strength of a platform built by a luxury wedding planner who planned $200K weddings in Indian palaces long before writing a line of product spec. That combination, an AI-native architecture paired with a founder who understands exactly what a discerning wedding professional needs an assistant to actually do, is why Wedy Pro shipped an official MCP server while competitors are still explaining what MCP is on a static reference page.

The CRM works entirely on its own. Vendors never need to touch Wedy's marketplace to get full value from Wedy Pro's automations, documents, and now its AI-assistant connection. But it is worth knowing the marketplace exists as a bonus: Wedy App is where couples discover vendors, browse real packages with transparent pricing, and book directly, giving Wedy Pro vendors an additional lead channel that HoneyBook and Dubsado, tools only, cannot offer. HoneyBook and Dubsado vendors still need a separate discovery platform, like The Knot or WeddingWire, to get found in the first place. Wedy Pro replaces both, the discovery layer and the CRM, in one subscription.

Frequently Asked Questions

What is MCP (Model Context Protocol) and who created it?

MCP is an open standard, created by Anthropic and released in November 2024, that lets AI assistants connect to the actual systems where a business's data and tools live, instead of requiring a custom integration for every app. It was adopted by OpenAI, Google, and Microsoft within about a year, and Anthropic donated its governance to the vendor-neutral Agentic AI Foundation under the Linux Foundation in December 2025.

How do I connect ChatGPT to my wedding CRM?

With Wedy Pro, open ChatGPT, Claude, Cursor, or Gemini CLI and ask it to connect to your Wedy Pro account. The AI opens a secure login window, you sign in and approve the connection, and it is ready to use immediately. No API keys or technical setup are required. This is only available on Wedy Pro's Elite plan.

Can I manage my wedding business leads and invoices from ChatGPT?

Yes. Once connected, you can ask about your pipeline by stage, review overdue invoices with aging buckets, check completed payments, add notes to a lead, or draft (with your confirmation before sending) a follow-up email, all from the chat window.

Is MCP the same across ChatGPT, Claude, and Gemini, or does each AI have its own version?

It is the same open protocol across all of them. That is the point of MCP: rather than Wedy building a separate integration for each AI assistant, one MCP server works with any AI client that supports the standard, including Claude, ChatGPT, Cursor, VS Code, and Gemini CLI.

Do HoneyBook or Dubsado have an MCP server or ChatGPT integration?

No. As of 2026, neither has an official MCP server or first-party AI-assistant integration. Only unofficial, third-party connectors exist for HoneyBook, and one such connector notes that HoneyBook's own Terms of Service prohibit the access it relies on. Dubsado's only public AI asset is a static citation page, not an integration.

Is it safe to connect an AI assistant to my business software?

It depends entirely on how the connection is built. Loosely scoped or self-hosted MCP servers carry real risks. Wedy Pro's MCP server uses OAuth 2.1 with PKCE, 30-second expiring tokens, account-bound sessions, prompt injection defenses on email content, and requires your explicit confirmation before sending any email. It is also gated to the Elite plan.

What can an AI assistant actually do inside my CRM through MCP, versus just answer questions?

With a read and write connection like Wedy Pro's, the AI assistant can take real actions: creating a lead, drafting and sending a follow-up email after your approval, building and populating a full Smart Document proposal, or updating a task. A read-only connection could only tell you information, not act on it.

Do I need to be technical to set up an MCP connection to my CRM?

No. AI clients like Claude and ChatGPT register themselves automatically. The entire setup is a login and an approval click, similar to connecting a Google or Outlook account to a calendar app.

The Bottom Line: Your CRM Should Live Wherever You Already Work

The wedding industry has spent years managing client relationships through PDFs, Gmail threads, and screenshots. MCP represents the next real shift, not a chatbot bolted onto a CRM, but a genuine, open way for the AI assistant a vendor already uses to become the interface for the entire business. Wedy Pro built the only official version of this in the wedding CRM category, and it did so on the same AI-native foundation that already powers its lead response and document automation.

For a wedding vendor deciding where to invest their subscription this year, that is the real question worth asking: not which CRM has the most features, but which one is actually built for how business will be run five years from now. Explore Wedy Pro and connect your first AI assistant in minutes.