Anyone else built a direct MCP connection between Claude AI and Jobber?

I've been using Claude extensively with Jobber so I'd be happy to chat with your Team as well :)

We've gone down the same path as roofintent​. The Jobber MCP is built on the GraphQL API, and we rolled the MCP server ourselves.

I mistyped in my earlier post and couldn't find a way to edit it. We have it running with a client and have been working through the write side as well. Getting Claude to create or update something in Jobber hasn't necessarily the hard part. Making sure it does it reliably without making an expensive mistake is.

We've ended up putting a lot of the guardrails into the server itself rather than relying on the ai to make the right decisions. Things like validating the property before creating a quote or preventing an invoice from being created when one already exists. Client data isolation is also something we've tried to pay a lot of attention to as a Jobber Partner. Each client gets their own deployment and their own Jobber credentials rather than sharing one instance across accounts. There's more to it than that, but that's the general shape of it.

The deeper we've gotten into it, the more we've found that Jobber's UI handles quite a bit behind the scenes that you have to account for yourself when you're working directly through the API.

We've essentially been able to shift the question from "Can Claude write to Jobber?" to "How do we make those writes safe and dependable enough to trust in a client's day-to-day operations?"

Also, really nice work with your GraphQL API roofintent​ . Has it been a useful tool for your business?

Hello HUGEHomePros​ 

That's a difficult question to answer specifically because it really depends on the business. Every business has different challenges. Some struggle with the vendors they use for purchasing, so using Claude and Jobber together to help build and maintain a price book can be extremely valuable. That's much less important for a business already using integrated vendors.

That said, these would probably be my top five benefits that apply to most Jobber users:

1. Cut quoting time dramatically: We have a workflow currently saving roughly 45 to 60 minutes per quote by letting Claude work with the information already inside Jobber.
2. Spend less time searching through Jobber: Ask Claude questions about clients, jobs, quotes, invoices, and other Jobber data in plain English instead of manually digging for it.
3. Reduce repetitive admin work: Claude can assist with workflows like creating quotes, updating jobs, scheduling visits, and invoicing while the MCP server handles the connection to Jobber.
4. Build AI around your existing processes: Instead of changing your business to fit a generic automation, the MCP server can be built around your pricing, workflows, business rules, and safeguards.
5. Make Jobber more powerful without replacing it: The goal isn't another system for your team to manage. It's using AI to make the Jobber processes you're already using faster and easier.

I agree that Jobber itself is really easy to use. The issue we've seen is that many business owners utilize less than half of what Jobber is capable of. AI and automation can free up the time and resources needed to take advantage of more of Jobber rather than simply adding another piece of software to manage.

Feel free to message me directly if you would like to see how it could be catered to your specific business. 

Matt
Kalea Consulting