MCP 服务器

CRE Intelligence MCP

Provides live commercial real estate data (rates, demographics) and analysis tools (DCF, rent roll parsing, lease abstraction, IC memo generation) within Claude Desktop.

README

CRE Intelligence MCP

Live market data for commercial real estate analysis — inside Claude.

Connect this MCP to Claude Desktop and instantly access Federal Reserve interest rates, Census Bureau demographics, DCF modeling, rent roll parsing, lease abstraction, and IC memo generation — all from a single prompt.

"Analyze this deal: 2201 South Blvd, Charlotte NC. NOI $400k, asking $6M, retail strip."

→ Pulls live SOFR from the Fed. Pulls Census demographics for that exact block. Builds a full levered 10-year DCF. Writes an institutional-quality IC memo. 30 seconds.

Why this exists

The #1 problem with AI in CRE: 66% of professionals use it daily, but only 5% trust it for actual deal decisions.

The reason? AI guesses at rates and demographics. A DCF built on a hallucinated SOFR rate is worthless.

This MCP fixes that. Every number comes from a verified public source:

Interest rates → Federal Reserve (FRED API)
Demographics → US Census Bureau ACS
Document analysis → Claude AI with structured output

Zero data licensing fees. Zero hallucinations on financial inputs.

Tools

Tool	What it does	Data source
`get_current_rates`	Live SOFR, 10yr Treasury, Fed Funds + implied cap ranges by property type	FRED
`get_inflation_data`	CPI, shelter inflation, rent CPI + DCF rent growth guidance	FRED
`get_cre_market_data`	CRE price index, C&I loan trends, delinquency rates, credit spreads	FRED
`get_market_demographics`	Median income, employment, vacancy, rents for any US address	Census Bureau
`get_radius_demographics`	1/3/5-mile trade-area rings: population, weighted income, renter share, rents	Census Bureau
`analyze_rent_roll`	Paste PDF text → structured JSON: tenant, SF, rent, dates, expirations	Claude AI
`abstract_lease`	Paste lease text → term, rent schedule, TI, options, red flags	Claude AI
`flag_lease_risks`	Rent roll JSON → rollover risk, concentration risk, due diligence checklist	Claude AI
`build_dcf_model`	Full levered 10-year DCF with live rates auto-fetched from FRED	Python + FRED
`generate_deal_memo`	Address + NOI + price → full IC memo with live rates and demographics	Claude AI + FRED + Census

Example output

Prompt: "Get me current interest rates"

SOFR:           3.63%   (June 8, 2026)
SOFR 30-day:    3.59%
10yr Treasury:  4.55%
5yr Treasury:   4.29%
Fed Funds:      3.63%

Implied cap rates (spread over 10yr T):
  Core Multifamily:  5.30% – 6.05%
  Core Industrial:   5.55% – 6.30%
  Core Office:       6.05% – 7.05%
  Value-Add:         6.05% – 7.05%

Loan rate guidance:
  Floating: SOFR + 150–250bps = ~5.38%–5.88%
  Fixed:    10yr T + 150–200bps = ~6.05%–6.55%

Prompt: "Analyze this deal: 2201 South Blvd Charlotte NC, NOI $400k, asking $6M, retail strip"

The MCP automatically chains get_current_rates + get_market_demographics + build_dcf_model + generate_deal_memo and returns a full IC memo including:

Entry Cap Rate:   6.67%  (+212bps over 10yr Treasury)
Loan Rate:        5.38%  (derived from live SOFR 3.63% + 175bps)
Year 1 DSCR:      1.53x
IRR:              14.6%
Equity Multiple:  3.11x

Census Tract demographics (2023 ACS):
  Median HHI:       $141,419
  Employment rate:  97.2%
  College educated: 64.9%
  Vacancy rate:     8.6%

Recommendation: GO — subject to rent roll review and comp analysis.

Setup

Option A — Hosted (fastest, no API keys)

Add this to your claude_desktop_config.json and restart Claude Desktop:

{
  "mcpServers": {
    "cre-intelligence": {
      "type": "streamable-http",
      "url": "https://cre-intelligence-mcp.onrender.com/mcp"
    }
  }
}

If your MCP client only supports stdio servers, use the mcp-remote bridge instead:

{
  "mcpServers": {
    "cre-intelligence": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://cre-intelligence-mcp.onrender.com/mcp"]
    }
  }
}

Free during beta. All data fetching runs server-side.

Option B — Self-hosted

Prerequisites

Claude Desktop
Python 3.11+
Free API keys (takes ~2 minutes total):
- FRED: fred.stlouisfed.org/docs/api/api_key.html
- Census: api.census.gov/data/key_signup.html
- Anthropic: console.anthropic.com

Install

git clone https://github.com/Zwondra/cre-intelligence-mcp
cd cre-intelligence-mcp
python3.11 -m venv venv
venv/bin/pip install -r requirements.txt
cp .env.example .env
# Add your API keys to .env

Connect to Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "cre-intelligence": {
      "command": "/path/to/cre-intelligence-mcp/venv/bin/python3",
      "args": ["/path/to/cre-intelligence-mcp/server.py"],
      "env": {
        "ANTHROPIC_API_KEY": "sk-ant-...",
        "FRED_API_KEY": "your_fred_key",
        "CENSUS_API_KEY": "your_census_key"
      }
    }
  }
}

Restart Claude Desktop. The tools will appear automatically.

Test it

Open Claude Desktop and say:

"Get me current interest rates"

You should see it call get_current_rates and return live Federal Reserve data.

Document analysis

The document tools (analyze_rent_roll, abstract_lease) work by pasting PDF text directly into the prompt. In Claude Desktop:

Open your rent roll or lease PDF
Copy all the text
Say: "Analyze this rent roll: [paste text]"

The tool extracts every tenant, suite, SF, rent, lease dates, and expiration into structured JSON — then flag_lease_risks can immediately analyze it for rollover and concentration risk.

Data sources

Source	What	Cost
FRED (Federal Reserve)	SOFR, Treasury yields, Fed Funds, CPI, CRE price index	Free
Census Bureau ACS	Income, employment, housing, rents by census tract	Free
Anthropic Claude	Document parsing, risk analysis, memo generation	Pay per use