XBert MCP - Your Trusted Layer
Model Context Protocol (MCP) is a secure, standard way for Claude, ChatGPT or whichever AI you use to connect to the software your practice already runs, and work with it on your behalf.



In Claude or ChatGPT's settings, add a custom connector with one address: mcp-gateway.xbert.io/mcp.
That's the whole technical bit.
A browser window opens and you sign in with your XBert account. Your assistant gets your permissions, no more, no less, and your credentials stay with XBert.
Ask it the way you'd ask a colleague. “Who owes us the most?” and “who's over capacity next week?” work better than anything technical ever will.
Ask it like you'd ask your best senior
Start with the work, not the technology. Every practice has the same Tuesday-morning questions — some about a client's books, some about how the practice itself is tracking.
One product, two surfaces
Everything your assistant does also lives as a screen inside XBert — same data, same rules, same audit trail. Chat for ad-hoc questions; the app for daily operations and team visibility.
Letting an assistant near the books is a big step. Every write is validated and checked for duplicates first, and marked as a change to confirm - your assistant stops and asks before it submits.
What XBert guarantees: your exact permissions, real validation, and an audit-log entry on every action.
★★★★★

Every Client, Every Job.
One Question Away.
Manage your practice with XBert. Connect MCP. No code, no API keys - just sign in and start asking.
What is the XBert MCP?
MCP — the Model Context Protocol — is a secure, standard way for an AI assistant like Claude or ChatGPT to connect to XBert and work with it on your behalf. You ask in plain English; the assistant uses XBert's tools to answer, and to draft work for your review.
Do I need to be technical to use it?
No. Setup is paste-one-address-and-sign-in, and from there you type questions the way you'd ask a colleague. There's no code, no API keys and no syntax to learn.
Which AI assistants work with it?
Claude on the web and Claude Desktop, Claude Code, and any other assistant that supports MCP connectors. ChatGPT works too, with a bit more setup — see the setup guide for what your plan and workspace admin need to allow.
Is my client data safe?
You sign in with your XBert account using OAuth, so your credentials stay with XBert and the assistant inherits your XBert role — it can only see what you can see. Write actions are marked for your assistant to confirm, and every action is written to the XBert audit log.
Can the assistant change my clients' books without me?
Not without it being put to you first. Write actions — creating a bill, recording a payment, updating a contact — are validated, checked for duplicates and marked as changes to confirm, which is what makes Claude and other MCP clients stop and ask you before submitting.
Which clients can my assistant work with?
Two things decide what your assistant can reach: your XBert role, and the client's plan. If you can't see a client in XBert, your assistant can't either — and the data and reporting tools only run for clients on an AI-enabled plan.
Do I need a new account or login?
No. You connect with your existing XBert login, and your existing XBert role and permissions carry over — there's nothing new to create. Not on XBert yet? Start at xbert.io.
How do I remove or manage access?
Remove or disable the XBert connector in your assistant's settings and the connection is gone. Reconnect any time by signing in again — access always follows your current XBert role, and every action leaves an audit trail.
Where do I go to connect?
Head to the XBert MCP connect page — it walks you through adding the connector and your first five prompts in about five minutes.
What can my assistant actually do with XBert?
Eleven areas covering both halves of the week: your clients' books (receivables, payables, bank, payroll, reporting) and your own practice (jobs, WIP, capacity, worklists) if you run Xero Practice Manager. See the full breakdown on the features page.