Property Intel MCP

Overview

The Property Intel MCP server provides AI agents with direct access to US property and rental data covering 158 million+ properties. It connects to ATTOM Data and RentCast APIs to deliver property characteristics, full multi-decade sales history, assessed and market values, ATTOM AVM (automated valuation), tax records, deed-holder names and mailing addresses, active mortgage details, rental listing status, rent estimates, and comparable rentals — all formatted for optimal AI consumption.

Key Capabilities

• search_properties: Search properties by zip code or state/county with rich filtering options. Returns property type, beds/baths, square footage, lot size, year built, last sale price and date, price per square foot, assessed value, market value, annual taxes, ATTOM AVM (current automated valuation with high/low confidence range, month-over-month change, and confidence score), and owner occupancy status. Supports filtering by property type (SFR, condo, apartment, commercial), bedroom count, year built, sale price range, and sale date range. Results are paginated at 20 per page.

• calculate_rental_yield: Calculate gross yield, net yield, and cash flow for any property. Provide an address to auto-pull the market value and real tax amount from ATTOM, or enter financials manually. Returns a full breakdown including gross/net yield percentages, monthly and annual cash flow, and an itemized expense summary. The address-mode call is backed by the same expanded-profile endpoint as the owner-details tool, so the underlying lookup also has the deed-holder and active-mortgage context available — no extra API hops needed when an investor wants to drill in.

• check_rental_status: Check if a property is currently listed for rent and get a rent estimate with comparable rentals from RentCast. Provide an address to see the listing status, AVM rent estimate with low/high range, and comparable rentals sorted by similarity — or search by zip code to survey area rental listings. Filter by bedrooms, bathrooms, or property type. Pairs naturally with calculate_rental_yield: get the rent estimate first, then calculate the full yield.

• get_property_owner_details: For a single property by address, returns the current deed-holder name(s), mailing address (with absentee indicator when it differs from the property address), LLC vs individual flag, owner-occupied status, and active mortgage details (loan amount, lender, originated date, term, interest rate, loan type code, maturity date). Surfaces up to four co-owners and second-position mortgages when present. The natural follow-up after search_properties when an investor wants to know who owns a specific property and how it is financed. Does NOT return phone numbers or email addresses — skip-trace data is intentionally out of scope.

• get_sale_history: For a single property by address, returns the full chronological sales and financing history on file (typically 5-15+ entries spanning 20+ years). Each entry includes the sale date, price (or non-disclosed flag), buyer name(s), seller name(s), deed type, title insurer, and the mortgage on that sale (loan amount, lender, term, rate, maturity). Useful for tracing the deed trail, detecting recent flips, analyzing length of ownership, and spotting lender or title-company churn. Sales returned in reverse chronological order with the current deed holder shown above the timeline for orientation.

Use Cases

• Real estate investors can run the full analysis pipeline: search for properties by location, check rental status and get rent estimates, then calculate yield with real tax data. Spot absentee owners, identify the deed holder and mailing address (LLC or individual), see the active mortgage on the property, pull the multi-decade deed trail to detect recent flips or long-tenured equity-rich owners, verify if they're renting, and model cash flow — all in one conversation.

• Property analysts can pull assessed values, market values, ATTOM AVM, and tax data for an entire zip code or county to build comps, spot undervalued properties, or analyze market trends. Cross-reference with rental comps to assess investment potential.

• Real estate agents can quickly look up recent sales in a client's target area, filtered by beds, property type, and budget, to prepare market reports. Check what comparable properties are renting for to advise investor clients. Pull a property's full deed trail to brief sellers on chain-of-title before listing.

• Portfolio managers can search across counties to identify absentee-owned properties, pull the deed holder and mailing address for outreach, check active mortgage details to find equity-rich or distressed positions, run sale history to find recently flipped or long-held parcels, verify rental status, track sale activity by date range, and monitor price movements. Run yield calculations across properties to rank by net return.

• Proptech developers can integrate property search, valuation, owner & mortgage lookup, sales-history analysis, and rental market capabilities into their AI-powered applications — chatbots, market analysis tools, or automated reporting systems — without building direct API integrations.

Who This Is For

This MCP server is built for AI agents and the developers who build them. Whether you're creating a real estate chatbot, an investment analysis tool, a market research assistant, or any application that needs structured US property and rental data, this server gives your AI direct access to ATTOM and RentCast data through a clean, well-documented tool interface.

Non-Disclosure State Support

In states like Indiana, Texas, and others where sale prices are not part of the public record, the server automatically detects this and clearly notes it in results. Assessed and market values from county tax records are shown instead (and the ATTOM AVM gives a current third estimate), so users always have valuation data regardless of disclosure rules. In get_sale_history, non-disclosed sales still surface buyer/seller/deed-type/lender — only the dollar amount is suppressed.

Data Coverage

ATTOM Data (property, sale, tax, owner, mortgage, AVM, sales history):

• 158 million+ US properties
• 99% US population coverage
• Data sourced from public records, tax assessors, recorded deeds, and mortgage records
• Multi-decade sales history (typically back to early 2000s where records exist)
• Active first- and second-position mortgages with lender, amount, term, rate, maturity
• ATTOM AVM (Automated Valuation Model) with confidence scoring
• Updated weekly

RentCast (rental listings, rent estimates):

• 140 million+ US property records
• ~96% residential listing coverage across all 50 states
• Active and inactive rental listing data
• AVM rent estimates with comparable properties
• Rental listing data updated daily