Connect Fran-Trak to your AI assistant

Claude (claude.ai or Claude Desktop)

1. Open Settings → Connectors and choose Add custom connector.
2. Name it Fran-Trak and enter https://mcp.fran-trak.com/mcp.
3. Click Connect, sign in with your Fran-Trak username and password, and allow access.

Charts and tables render right in the chat when you ask for sales, reorder suggestions, current status or an order. Requires a paid Claude plan.

ChatGPT

1. Open Settings → Connectors (Developer mode may need to be enabled by your ChatGPT admin) and choose Create.
2. Enter https://mcp.fran-trak.com/mcp with authentication set to OAuth; leave the Client ID blank.
3. Sign in with your Fran-Trak username and password.

Google Antigravity (desktop app, IDE or command line)

Antigravity shares one setup across the desktop app, the IDE and the agy command line, so do this once.

From the command line:

agy mcp add fran-trak https://mcp.fran-trak.com/mcp
agy
> /mcp

From the desktop app or IDE:

1. Open Agent Settings (Cmd+, on Mac, Ctrl+, on Windows) → Customizations → Manage MCP Servers → View raw config, and add:

   { "mcpServers": { "fran-trak": { "serverUrl": "https://mcp.fran-trak.com/mcp" } } }

2. Back in Customizations, click Authenticate next to fran-trak.
3. Sign in with your Fran-Trak username and password, copy the code Antigravity shows you, paste it into the settings panel and click Submit.

The sign-in is shared by all three Antigravity surfaces and renews itself. Gemini CLI was retired by Google in June 2026; if you still have the old gemini command, use Antigravity instead.

Claude Code

claude mcp add --transport http fran-trak https://mcp.fran-trak.com/mcp
claude
> /mcp

Choose fran-trak in the /mcp menu to sign in.

Interactive views

Five kinds of answer arrive as a live panel in the chat instead of text. Each is opened by a particular question; you can then change dates, locations or sorting inside the panel and the assistant follows along.

• Sales dashboardKPI tiles and a chart. How were sales last week? · Compare the downtown and airport stores by week this month.
• Report tableAny catalogue report, sortable and exportable. Run the daily sales report for September. · Show me the labor detail report for both stores this pay period.
• Current statusToday so far, per store. How are we doing right now? · Is today's data current for every store?
• Reorder suggestionsProducts at or below par, with suggested quantities. What should the downtown store reorder?
• Order cardOne online order with its actions. Show me order 4821. · Open the next pending online order.

Sales KPIs, alerts, top sellers and inventory questions that don't match one of these come back as text.

What you can ask

"How did we do last week compared to last year?" · "Any alerts today?" · "What should the downtown store reorder?" · "Show me the labor detail report for this pay period" · "Which online orders are waiting?"

The assistant can look up:

• Sales, transactions and average check, by day, week, month or hour.
• Your best and worst sellers.
• Today so far, store by store.
• Your dashboard alerts, and which ones are red.
• Many of the portal's reports.
• Stock on hand and what to reorder.
• Online orders and their status.

It can also make changes you ask for:

• Mark an online order ready, send it to the register again, or cancel it.
• Refund an online order's card payment.
• Send a message to an online order's customer.
• Set a product's par level.
• Change a menu item's price at your stores.
• Mark menu items out of stock when you run out, or when a product runs out, and back in stock later.

Everything else only reads your data. Most AI apps ask you to approve an action before it runs. Check what you are approving. Questions: contact Fortibus support.