Skip to content
Works with Claude · Claude Code · ChatGPT · Microsoft Copilot

Connected to Claude in about five minutes.

No code, no API keys, nothing to install on a server. You paste one address into Claude, sign in with your XBert account, and start asking — about a client's books, or about how the practice itself is tracking. ChatGPT and Microsoft Copilot take more setting up, and what your plan and your IT admin have to allow is below. On Claude, most people get their first real answer before the kettle's boiled.

Under the hood: a remote MCP server with OAuth. That's the whole trick.

Full illustrated guide with screenshots: the XBert setup guide →

Before you begin, you need three things.

01

An XBert account

With your clients' files connected as usual — Xero, QuickBooks, MYOB or FreeAgent, plus Xero Practice Manager if you use it: XPM is what powers the practice-side questions about jobs, timesheets and WIP. The MCP works with what's already in XBert. One thing worth knowing: questions across all your clients cover the ones on an AI-enabled plan — Financial AI or Workflow AI — and the assistant tells you which it left out. Questions about the books need the client's ledger connected, and a client on hold for an unpaid invoice can't be read at all. Not on XBert yet? Start at xbert.io.

02

Claude, or another AI that speaks MCP

Claude is the straightforward path — web, Desktop or Claude Code, and you paste one address. ChatGPT takes more setting up: custom MCP servers are added on the web only, and on a Business, Enterprise or Edu workspace an admin has to allow them and publish XBert before anyone else can use it. Microsoft Copilot is a different shape again — you reach XBert by building a small agent in Copilot Studio, and someone in IT approves it before the rest of the practice sees it. Each one has its own steps below.

03

Your normal login

That's the security model in one line: the assistant signs in as you and can only reach the clients you can reach in XBert. If you can't see a client in XBert, your assistant can't either. Access isn't the only gate, though — the client's plan decides what it can answer. On a workspace plan, connector settings are usually an admin's to switch on — ChatGPT in particular needs one to allow custom MCP servers before you can add anything.

Connect your assistant.

One address, whichever assistant you use. Pick yours below — each set of steps has its own link, so you can send a colleague straight to the right one.

XBert MCP addresshttps://mcp-gateway.xbert.io/mcp

The straightforward path. Add it once and XBert is there everywhere you use Claude, with no extra setup on your other devices.

The first button opens Claude’s Add custom connector box with the name and address already filled in — sign in to Claude first and it lands ready to add. Both buttons end up on the same screen, so use the second if you would rather type it in yourself.

  1. Select Customize in Claude’s left sidebar, then the Connectors tab.
  2. Select + Add, then Add custom connector.
  3. Name it XBert, paste the address above as the server URL, and select Continue. Keep what Claude picks on the next screen — Register automatically for the OAuth client — and select Add. XBert registers itself, so there is no client ID or secret to fetch from us; if your screen shows Client ID and Client Secret boxes, leave them empty.
  4. Sign in with your XBert account in the window that opens. If it doesn’t open, select Connect next to XBert.
  5. Back in the chat, XBert now appears in your tools menu. Ask away.

On a Team or Enterprise plan? An Owner goes first.

Members can’t add a custom connector on those plans. An Owner or Primary Owner — on Enterprise, also anyone whose custom role includes Manage access to Libraries — adds XBert once for the whole firm under Organization settings → Connectors → Add → Custom → Web, and everyone else then finds XBert under Customize → Connectors and selects Connect. On Free, Pro and Max you add it yourself, with no admin involved — though Free allows only one custom connector at a time.

Step-by-step with screenshots in the XBert setup guide →. Claude’s own version: adding a custom connector →

If you have already added XBert on claude.ai, you may have nothing left to do.

Try this first. Connectors you have added on claude.ai load automatically in Claude Code when you are signed in with that same Claude account. Run /mcp and look for XBert before you add anything. This only applies to a Claude subscription login — not an API key.

If it isn’t there, or you would rather register it locally, run:

claude mcp add --transport http --callback-port 6274 xbert https://mcp-gateway.xbert.io/mcp
  1. In your next session, run /mcp and complete the XBert sign-in when prompted.
  2. Ask in plain English.

--callback-port 6274 is not optional. XBert registers a fixed port to finish the sign-in, and Claude Code otherwise picks one at random — leave the flag out and the sign-in fails. --scope user makes XBert available from any folder rather than only the one you ran the command in.

If your Claude Code doesn’t recognise --callback-port, update it first — the flag is only in recent versions. Full syntax in Claude Code’s MCP documentation →

ChatGPT adds custom MCP servers on the web, not in the mobile apps. On a Business, Enterprise or Edu workspace an admin has to allow them first.

ChatGPT has no one-click install, so the first button only opens the page where you add the connection — you still paste the address yourself.

On a Business, Enterprise or Edu workspace, an admin goes first

An admin or owner opens Workspace settings and, under Permissions & roles, turns on Create plugins with MCPs (older workspaces show this under Connected Data). On Business, only an admin or owner can add XBert, and they then publish it for everyone else.

  1. Go to Plugins, select +, then Add custom MCP server.
  2. Name it XBert with a short description, and under Connection paste the address above as the server URL — keep the /mcp on the end.
  3. Choose OAuth for authentication, read the risk warning and select I understand and want to continue, then Create as a plugin. Sign in with your XBert account when the window opens.
  4. On a workspace, an admin or owner shares it with everyone else from Plugins → Personal, using the ••• menu and Publish.

No Add custom MCP server option? Some accounts still need Developer mode switched on first, under Settings → Plugins → Advanced settings (older workspaces: Settings → Apps → Advanced settings).

Check what your own plan allows. OpenAI documents this for Business, Enterprise and Edu, where write actions are in beta, and read-only access on Pro — there, XBert can answer questions but not take action. Plus isn’t documented. OpenAI has called this area Connectors, Apps and now Plugins, and its own help pages don’t all agree yet, so if the wording differs, look for wherever custom MCP servers are added. OpenAI on MCP in ChatGPT →

Copilot Studio recognises XBert's sign-in on its own — there is no client ID or secret to copy from us. But “Copilot” means five different Microsoft products, and only one of them is a route to XBert.

Which Microsoft Copilot products can reach XBert
ProductCan it reach XBert?
Copilot StudioMicrosoft's agent builder.YesThe one documented route. You add XBert as a tool on an agent, and it recognises our sign-in on its own.
Microsoft 365 CopilotCopilot Chat in Teams, Word, Outlook and on the web.IndirectlyYou can't paste our address in here. Microsoft now supports MCP servers through a packaged plugin or declarative agent, but that needs a package XBert hasn't published yet — so today you reach it through a Copilot Studio agent added to Teams and Microsoft Copilot.
Create agent / Agent BuilderThe lightweight builder inside Microsoft 365 Copilot.NoTakes knowledge sources only. Microsoft's own advice is to use Copilot Studio when you need to reach an outside service.
Microsoft Copilot (consumer)copilot.microsoft.com, and the Copilot mobile app.NoOffers a fixed list of connectors — OneDrive, Outlook, Gmail and the like. There is no way to add your own.
GitHub CopilotThe coding assistant in VS Code and Visual Studio.NoA different product entirely — and not the same thing as the “GitHub Copilot harness” that runs Copilot Studio's newer experience. Listed only so nobody goes looking for XBert in the wrong place.

Microsoft publishes no one-click install link for adding a tool, so this one is click-steps rather than a button. Copilot Studio now opens in its newer experience (in preview), where the path is Build → Tools → Add → Model Context Protocol (MCP). Microsoft documents the sign-in steps below for classic agents: to build one, select … at the bottom of the home page’s side bar and then Open classic experience, or choose Other ways to build when you create the agent. Microsoft on switching →

Three things to sort out first.

  • Generative orchestration must be on. In the agent, that is Settings → Generative AI. New agents have it on already. Without it, MCP tools don’t run at all.
  • Your IT administrator may need to act. MCP servers in Copilot Studio run over Power Platform connectors, so a data policy that blocks those also blocks XBert — quietly. Send your admin to the Power Platform admin centre (Security → Data and privacy → Data policy), or to Microsoft’s write-up →
  • Licences. Whoever builds the agent needs a Copilot Studio or Microsoft 365 Copilot licence, and your organisation needs Copilot Credits — a prepaid pack or pay-as-you-go. People who only use the finished agent don’t need a licence. Worth knowing: the free trial can build and test an agent but cannot publish it. Microsoft’s licensing guide →
  1. Open your agent in Copilot Studio and go to its Tools page.
  2. Select Add a tool → New tool → Model Context Protocol.
  3. Server name: XBert. Server description: a sentence or two on what XBert does — Copilot reads this to decide when to call XBert, so it is worth writing properly. Server URL: the address above.
  4. For authentication choose OAuth 2.0, then Dynamic discovery. You do not need a client ID, a client secret, or any URLs from XBert — that is the whole point of this option. Don’t choose Manual: we don’t issue the credentials it asks for.
  5. Select Create, then Create a new connection, sign in with your XBert account, and select Add to agent.

Leave all seven tools on. XBert shows Copilot Studio just seven tools: three that find and run the rest of XBert on demand, and four for signing in and choosing your business. That keeps you well inside Microsoft’s limit, and new XBert capabilities reach your agent without you doing anything. Leave Allow all on — switch any of the seven off and the agent can lose its way into XBert. Questions? Ask us.

Getting it in front of the rest of the practice.

Publish the agent, then under Channels add Teams and Microsoft Copilot. You can install it yourself and share its install link with named colleagues; listing it for the whole practice needs an administrator’s approval. Test it in the Test agent panel or in Teams — sign-in doesn’t work on the Demo Website channel, so that one will look broken when it isn’t. Microsoft on agent approval →

Anything that speaks MCP can point at the same address. Whether it can finish signing in is the part worth checking first.

Goose

Goose is the one desktop assistant besides Claude Code that can complete our sign-in, because it lets you set the port it comes back on. Set GOOSE_OAUTH_CALLBACK_PORT to 6274 first — the sign-in will not finish without it.

  1. For anything else: find where your assistant manages connectors or MCP servers, and add a remote MCP server with the address above.
  2. If it asks for a transport type, choose HTTP (sometimes shown as “streamable HTTP”). If it asks for an OAuth client ID, leave it empty — XBert registers itself.
  3. Complete the XBert sign-in when your browser opens, then start asking.

Assistants we can’t connect yet, and why.

We only publish a button where we have checked the whole journey — install, sign in, first answer. These ones install fine and then can’t finish signing in, so we’d rather tell you than let you find out.

Assistants that cannot complete XBert sign-in today
Visual Studio CodeSigns in on a fixed port of its own (33418), with no way to change it.
CursorThe desktop app signs in on a fixed port of its own (8787); its web and Agents sign-in uses a Cursor address we don't accept yet.
Devin Desktop (formerly Windsurf)Supports sign-in, but doesn't document which port it comes back on, so we can't say it will finish.
ZedSigns in on a different port each time, with no way to fix it.
JetBrains IDEsAI Assistant has no sign-in flow for remote servers yet, and Junie doesn't document the port it signs in on.
LM StudioSigns in on a fixed port of its own (33389), with no way to change it.
ClineSigns in on fixed ports of its own (1456 to 1461), with no way to change them.

Using one of these? Tell us which — it helps us decide what to support next.

Your first five prompts.

Start read-only — a couple about a client's books, a couple about the practice itself. Get a feel for how it answers before you ask it to do anything.

“What can you do with my XBert account?”
“Which of my clients can you see in XBert?”
“Summarise this week's XBert alerts for [client].”
“Show me aged receivables for [client] — who should I chase first?”
“What's on my worklist today, and which jobs are running behind?”

Name the client and the period — or the person and the week. “Bayside Cafe, last quarter” beats “that cafe file” every time.

Managing and removing access.

Turn it off in one place.

Remove or disable the XBert connector in your assistant's settings and it stops connecting. Reconnect any time by signing in again. To end someone's access for certain — a lost laptop, someone leaving — remove them in XBert: every call is checked against XBert, so their assistant loses access with them.

Access follows your client access.

The assistant only ever holds the permissions of the person signed in. Change someone's client access in XBert and their assistant's reach changes with it — nothing is granted beyond that.

Everything leaves a trail.

Every change the assistant can make is marked as a change to confirm, so your assistant stops and asks before it submits. That approval prompt belongs to the assistant rather than to XBert — what XBert guarantees is your permissions, the validation and an audit-log entry on every action.

Sign-in uses OAuth. Your credentials stay with XBert — the assistant never sees your password.

The technical bit, for whoever asks.

Most people never need this. Your IT team might, and it saves a round trip if it's written down.

How the connection works

Address
https://mcp-gateway.xbert.io/mcp
Transport
Streamable HTTP
Sign-in
OAuth 2.1 authorisation code, PKCE (S256) required
Client registration
Dynamic — nothing to obtain from XBert
Scopes
openid, profile, email, offline_access

XBert is reached over the public internet, so there is no firewall change to make for the assistants that run in the cloud. If your firm manages your devices and filters outbound traffic, ask IT to allow mcp-gateway.xbert.io, auth.xbert.io and cdn.auth0.com — the sign-in page loads from the last two. If you sign in to XBert with Google, Xero or QuickBooks, their sign-in pages need to be reachable too.

Known limits, stated plainly

  • Assistants on your own machine sign in on one fixed port. XBert accepts 6274 and nothing else. Claude Code and Goose let you set it; several others don’t, which is why they’re on the list above rather than behind a button.
  • A client’s plan decides what can be answered. Questions across all your clients cover the ones on an AI-enabled plan — Financial AI or Workflow AI — and the assistant says which it left out. Questions about the books need the client’s ledger connected. A client on hold for an unpaid invoice can’t be read at all.
  • It doesn’t post to the ledger. Your assistant reads your clients’ books and works on the practice side — tasks, notes, comments, schedules — but it doesn’t create or change transactions in Xero, QuickBooks, MYOB or FreeAgent.
  • Your assistant does the asking, not us. The prompt you see before a change is submitted belongs to the assistant. What XBert guarantees is your permissions, the validation, and an audit-log entry on every action.

Assistant menus move constantly. The steps above were checked against each vendor’s own documentation on 12 October 2026 — if a menu name has changed since, look for the nearest equivalent, and tell us so we can correct this page.

How to work with it well.

  • Start with questions, not changes.

    Spend your first week asking — balances, alerts, worklists. Move to creating and updating tasks once the answers have earned your trust.

  • Be specific.

    Client, period, and what good looks like. Specific asks get specific answers.

  • One client before twenty.

    Prove a routine on a single file, then let the assistant repeat it across the portfolio.

  • Review anything client-facing yourself.

    Always. The assistant drafts; you decide what leaves the practice.

  • Treat it like a bright new hire.

    Clear instructions, check the work early on. Trust builds the same way it does with people — quickly, when the work keeps standing up.

Connected? Here's where it gets fun.