Your AI assistant can now run Resideline

Resideline now ships an MCP server. MCP is the Model Context Protocol, the open standard that lets assistants like Claude, ChatGPT, and Cursor call real tools instead of guessing from training data. Connect it once and your assistant can run a full Resideline analysis on any US address, mid-conversation.

The workflow

Here is what it looks like in practice. You are reading a wholesaler blast. You copy the address and ask Claude:

"Run 123 Main St, Tampa, FL 33634 through Resideline and tell me if it is a deal at 250k."

Your assistant calls the Resideline tools, waits for the report, and comes back with the market value, the sold comps behind it, the rent estimate, and its own read on your 250k basis. No tab switching. No copying numbers between windows. The analysis happens where you were already working.

And because the tools return real comp data, the follow-up questions work too. Ask why the value is what it is. Ask which comp is doing the most work. Ask what rent does to the deal at 20 percent down. The numbers are in the conversation, so the conversation can use them.

The tools

The server exposes four tools:

• •analyze_property: the full report. Market value, sold comps, rent estimate, and a photo-verified condition grade. Uses one report from your plan.
• •get_property_value: a fast sale-value estimate with a value range, confidence, and top comps.
• •get_rent_estimate: long-term market rent with a range, projected yearly revenue, and rental comps.
• •get_accuracy_scoreboard: the live graded accuracy stats. Free, and it needs no API key.

Setup takes two minutes

First, get your API key. Log in at dashboard.resideline.com and copy it from your Profile page. Then add the server to your client. For Claude Desktop, that is one block in claude_desktop_config.json:

{
  "mcpServers": {
    "resideline": {
      "command": "npx",
      "args": [
        "-y", "mcp-remote", "https://resideline.com/mcp",
        "--header", "Authorization: Bearer YOUR_API_KEY"
      ]
    }
  }
}

Claude Code and Cursor configs are on the setup page at resideline.com/mcp-server. Paste, restart your client, and ask it to run an address.

The accuracy receipts come along

This is the part we care about most. Any tool can hand your assistant a number. The question that matters is whether the number is any good, and most valuation tools will not answer it.

We do, in public. Every Resideline estimate is frozen while the home is still for sale, then graded against the actual closing price after it sells. Right now the scoreboard shows 2.79% median error across 2,455+ graded closings, built from 356,000+ frozen predictions across 50 states.

The get_accuracy_scoreboard tool puts those live numbers directly in your assistant's hands, free, with no API key. So when your assistant tells you a property is worth 280k, you can ask it to cite the track record behind that number, and it can. If you want the full argument for why a public error rate is the only claim worth trusting, read why accuracy receipts matter.

Try it on your next deal

Grab your API key from the dashboard, paste the config from resideline.com/mcp-server, and forward the next address you are considering to your assistant. MCP calls use your plan's monthly reports, and the free plan includes 3 reports a month, so you can test the whole loop without paying anything.