AgencyGraph MCP Recipes: Let Your AI Assistant Answer Questions About Estate Agencies

CASAFARI operates as the property data layer for estate agents and their AI assistants, covering about 41,000 agencies and offices with homes for sale in 4,767 Spanish towns and about 17,600 in 309 Portuguese towns as of the October 2026 weekly build (about 22,000 in Spain and 6,600 in Portugal have 5 or more active properties and receive full benchmarking measures). You can explore real outputs in the sample reports in English or review all tools and pricing on the capabilities page.

How to connect Claude or Cursor to the AgencyGraph MCP server

Connecting to the AgencyGraph MCP server (https://platform.casafari.com/mcp/agencygraph) enables Claude, Cursor and other MCP clients to query a digital inventory index of Spanish and Portuguese estate agencies. One browser-based OAuth sign-in handles authentication. In Claude, navigate to Settings → Connectors → Add custom connector, paste the server address, and sign in with your CASAFARI account. You can also add https://platform.casafari.com/mcp/locations to resolve place names into canonical location IDs using the free locations_search tool. For automated pipelines, every MCP tool has a corresponding REST endpoint on the CASAFARI platform. Start with 1,000 free tokens for 30 days at https://platform.casafari.com/signin.

Recipe 1: How do you benchmark the estate agencies of a town?

Benchmarking provides objective figures on portfolio value, growth, price changes, overpricing at start, and time on the market. Prompt your AI assistant:

"Benchmark the estate agencies operating in Málaga by portfolio value and time on market."

1. locations_search(query="Málaga", country="ES") — 0 tokens (resolves the town to its ID).
2. agencygraph_area_benchmark(country="ES", location_id="...", sort_by="portfolio_value") — 75 tokens per page of 25 agencies (€3).

This workflow serves agency owners evaluating market position and investors, developers and banks choosing agency partners (see the Málaga sample report in English and Barcelona selection sample report in English). All figures reflect asking prices, never closing prices.

Recipe 2: How do you profile one estate agency by its name or website?

Map a trading name or domain to an agency profile showing active homes and territory metrics:

"Give me a performance profile and town footprint for Example Imobiliária in Portugal."

1. agencygraph_resolve_agency(country="PT", query="example-imobiliaria.pt") — 0 tokens.
2. agencygraph_agency_profile(country="PT", agency_id="...") — 50 tokens (€2).
3. agencygraph_territory(country="PT", agency_id="...") — 50 tokens per page of 25 towns.

This firm-level profile aids agency owners pitching for mandates (see the Cascais instruction sample report in English).

Recipe 3: How do you screen agencies for recruitment or prospecting?

Filter thousands of agencies by operational size and growth metrics:

"Find independent estate agencies in Lisbon and Porto with between 20 and 80 properties for sale."

1. locations_search for target towns — 0 tokens.
2. agencygraph_screen_agencies(country="PT", location_ids=[...], independent=true, min_properties=20, max_properties=80) — 100 tokens per page of 25 matches (€4).

This dataset assists franchise networks expanding their footprint and CRM vendors prospecting (as shown in the Lisbon and Porto vendor prospect list sample report in English).

Recipe 4: How do you track a watchlist of competitors every week?

Monitor changes in instruction volume and asking prices across a stored list of agency IDs:

"Show me weekly instruction changes and price adjustments for my saved watchlist of agencies."

1. agencygraph_resolve_agency (run once to store canonical IDs) — 0 tokens.
2. agencygraph_changes(country="PT", agency_ids=[...], since="2026-09-01") — 25 tokens per call for up to 25 agencies.

Recipe 5: How do you measure a franchise network against a town's market?

Assess network coverage and market share against local benchmarks:

"Evaluate the market share and office network of a franchise brand across Portuguese towns."

1. agencygraph_network(country="PT", brand="...") — 75 tokens per page of 25 offices.
2. agencygraph_market_share(country="PT", location_id="...") — 50 tokens per page of 25.

Network directors use this to identify coverage gaps (see the five Portuguese towns network gap sample report in English and five Spanish cities recruitment sample report in English). Property portals measuring agency coverage rely on similar queries. Note that town market shares can sum to over 100% because homes advertised by multiple agencies are counted once per participant.

Coming measures and AI visibility context

The following measures are coming to AgencyGraph, at no extra price; no dates promised: tech stack (CRM and website vendor), MLS membership, specialisation (luxury, new builds, rentals, commercial, foreign buyers), exclusivity, portal mix, SEO, AEO and GEO scores, and online presence.

In 150 Spanish and Portuguese towns, only about 7–13% of the sources AI assistants cite are an agency's own website; about 35% are portals (CASAFARI GEO study, October 2026, 1,199 questions).

To build your first agent, visit the AgencyGraph product page, inspect the AgencyGraph MCP server, submit data updates via the corrections page, or consult the technical reference on developers.casafari.com.