Sign inTry for freeRequest a demo

MCP Connector Reference

Connect AI assistants
to real estate intelligence

Complete reference for tools, skills, and data available through the RealAI MCP connector. Access 5.3 trillion real estate data points and comprehensive analytical AI workflows for Claude and other AI assistants.

Three layers of capability

The RealAI connector provides access to comprehensive real estate data and analytical workflows through three integrated layers.

MCP Tools

Low-level data access via the Model Context Protocol. Search, explore, and query structured real estate data.

Skills

High-level analytical workflows for specific real estate tasks like investment analysis, valuations, and market reports.

Data Catalog

Entities, topics, and fields available for querying. From national metrics to individual property records.

MCP Tools

Three tools are exposed through the MCP server. They are designed to be called in sequence: search → explore → query.

entity_search_by_name

Search for real estate geographies and multifamily properties using fuzzy name matching. This is typically the first tool to use when you have a location or property name but need its unique identifier.

Parameters

ParameterTypeRequiredDescription
querystringYesSearch term (min 2 chars). Use 1-2 words for best results.
limitnumberNoMax results (default: 10, max: 20)
typesarrayNoFilter to specific entity types
statestringNoTwo-letter state abbreviation

Filter Operations

Available filter operations for query_data. Use with the filters parameter.

EQUALSExact match
EQUAL_IGNORE_CASECase-insensitive match
INMatch any value in list
NOT_INExclude values in list
GREATER_THANNumeric greater than
GREATER_OR_EQUALNumeric greater or equal
LESS_THANNumeric less than
LESS_OR_EQUALNumeric less or equal
IS_NULLField is null
IS_NOT_NULLField has value
CONTAINSString contains
CONTAINS_IGNORE_CASECase-insensitive contains

Analytical Workflows

Skills

Skills are high-level analytical workflows that orchestrate tool calls and produce structured deliverables. The router skill automatically dispatches to the appropriate skill based on user intent.

Skill Invocation

Skills are invoked automatically based on intent. Use natural language describing your goal — the router handles dispatch.

Foundation skills for data access and routing

realai-routerAutomatic on real estate topics

Routes intent to the appropriate skill

realai-data"find", "lookup", "query", "what data"

Entity resolution, schema discovery, query construction

Data Catalog

The data catalog is organized around three concepts: entities, topics, and fields. Entities are the primary subjects you query data about. Each entity has identity fields plus access to various topics.

Geographic Entities

From national aggregates down to 100-household micro-areas

nationstatemarketsubmarketcountycensus_placezipcodeneighborhoodcentum

Property Entities

126M+ property records with multifamily and residential variants

propertyproperty_mfrproperty_residential

Reference Entities

Benchmark rates and economic indicators

mortgage_ratesnational_metrics_dailynational_metrics_monthlynational_metrics_quarterly

Available Topics

Topics are thematic groupings of related data. Different entities support different topics. Always use explore_data to discover exact names.

Geographic Topics

  • household_financials_snapshot — Income, net worth tiers
  • demographic_basics — Population, age, household size
  • employment — Employment rates, job growth
  • education — Educational attainment
  • credit_profile_snapshot — Credit scores, debt levels
  • migration — In/out migration patterns

Rent & Occupancy Topics

  • mf_rent_ts — Multifamily rent time series
  • mf_rent_ts_by_beds — Rent by bedroom count
  • sfr_rent_ts — Single-family rent time series
  • rent_roll_latest — Current rent roll snapshot

Property Topics

  • property_attributes — Building details, unit counts, vintage
  • mf_rent_and_occupancy_snapshot — Current rent/occupancy
  • tax_and_assessment — Property taxes and assessments
  • sales_history — Transaction history
  • ownership_and_management — Owner and manager info

Error Handling

Access denied: read:mcp scope required
Missing authenticationVerify user has proper access
Query parameter is required
Empty search queryProvide search term with 2+ characters
Entity is required for topics scope
Missing entity parameterInclude entity type when exploring topics
Limit cannot exceed 1000
Pagination limit too highReduce limit and use offset

Authentication

The RealAI Connector requires OAuth 2.0 authentication.

Rate Limits

The MCP server enforces rate limits per user and per organization. If you encounter rate limiting, reduce request frequency or contact your administrator.

Questions?

If you have questions about the RealAI MCP connector or need help with integration, please contact our team.