MCP server

MTR Analytics MCP

Connect a Pro account to Cursor, Claude, ChatGPT, or any client that speaks remote MCP. The agent can underwrite an address, compare homes for sale, find lease arbitrage, read your saved deals, and pull furnished monthly rates with Census and employment context.

For what this dataset is and why a chat should use it, read the guide.

Install

Remote MCP over Streamable HTTP. OAuth 2.1 with PKCE. Do not put a key in the URL.

https://mcp.mtranalytics.com/mcp

Cursor

Cursor Settings, then MCP, then add a server named MTR Analytics. Or paste this into ~/.cursor/mcp.json.

{
  "mcpServers": {
    "MTR Analytics": {
      "url": "https://mcp.mtranalytics.com/mcp"
    }
  }
}

Claude

In Claude, add a custom connector and use the server URL. In Claude Code:

claude mcp add --transport http "MTR Analytics" https://mcp.mtranalytics.com/mcp

ChatGPT

Add a custom connector. Server URL above. Authentication: OAuth. ChatGPT opens the browser for you.

Other clients

Use the server URL with HTTP transport and OAuth. If the client can only send a header, create a key in Settings after you are on Pro and send it as Authorization: Bearer. Browser sign-in is the path to use when the client supports it.

What it can do

Ask in your own words. One question can mix a listing, the comps behind it, the county's jobs, and the deals already in your account. The prompts below are examples of that mix, not a menu.

Underwrite a property

Give it a street address and an offer price. It runs the same Property Analyzer as the product: bedroom-matched furnished monthly comps, long-term rent, cash flow, cap rate, cash-on-cash, and DSCR, with Census and county employment for that place when they are published.

Underwrite 1847 E 2700 S, Salt Lake City, UT 84106 at $489,000 as a 3-bedroom furnished monthly rental with 25% down. Include the county employment picture.

Compare homes for sale

Name a city and a price range, and the monthly income you want. It pulls active listings and scores each one from the furnished monthly comps for that bedroom count, then you can underwrite the finalists.

I'm in Salt Lake City. Find active homes between $400,000 and $550,000 that should clear about $3,200 a month on furnished monthly comps, and tell me which ones are worth underwriting.

Find arbitrage leases

It lines up unfurnished asking rents against the furnished monthly revenue for the same bedroom count, and drops rooms in a shared house. The spread is a screen, not a finished cash-flow statement.

Find whole-home leases in Austin under $2,200 a month where the furnished monthly revenue still covers the rent after occupancy.

Compare your pipeline

It reads the deals saved in your account and returns the underwriting already attached to each one, ordered by monthly cash flow, so you can compare them without retyping the addresses.

Compare the deals in my pipeline. Which have the strongest cash flow, and which have been saved but not underwritten?

Hospitals and opportunities

One question can combine the Market Finder filters: rent, occupancy, income, miles to a hospital, and Level I or Level II trauma centers, plus saved local drivers such as a university, a military base, or a major employer. You can also list the hospitals in a city with their trauma level. When an Opportunities report is already saved, the agent can read it. It does not start a new one. A place with no saved university record is left out of that filter. That is not proof the place has no campus.

Find ZIP codes anywhere in the U.S. with at least two Level I trauma centers, a university, and an average monthly rate around $2,500.

Read a market

Rates, occupancy, revenue, comps, bedroom mix, monthly trend, metro rankings, and this month's market picks. Census income, renter share, and home values, plus county jobs and wages, come back with the place when we have them.

How does a 2-bedroom furnished monthly rental look in 78704, and do the county's jobs and the census renter share support that demand? Then rank Texas metros by monthly revenue.

Reading a number

Average monthly rate: Asking rent for a furnished stay of 30 nights or more. Not a nightly rate.
Monthly revenue: That rate times occupancy. Not cash flow after expenses.
Cash flow, cap rate, cash-on-cash, DSCR: A Property Analyzer underwrite for one address, with debt and expenses.
Screen yield: Annual comp revenue divided by a list price. A way to rank homes for sale, not a deal return.
Gross yield: Annual mid-term revenue divided by the median home value. A way to rank markets.
Arbitrage spread: Furnished monthly revenue minus the unfurnished asking rent. Utilities and furnishing are still open.
Census and employment: ACS demographics for the place, and BLS jobs and wages for the county. Median gross rent here is the Census rent, not the furnished rate.

Help

Allow says Pro is required.: Upgrade, then start the connect flow again.
The app returns and still shows signed out.: Remove the server and add the URL again.
An answer has no numbers.: Name the city and state, or the ZIP, in the same message. For one house, include the street address.
A listing you saw this morning is missing.: Market rates and comps are the latest published month. For-sale and lease searches use listings cached for about a day. Underwriting an address does not add it to your pipeline. Saving a deal still happens in the product.

Create an account · Pro pricing