Zillow Leads Property Data

What zillow-leads-property-data does

The zillow-leads-property-data MCP server gives an MCP-compatible agent access to the Zillow Leads & Property Data Apify actor. The actor returns Zillow listing records and can add agent and broker contact information, MLS attribution, price-history timelines, tax history, foreclosure or distress signals, school information, HOA fees, and property details such as heating, cooling, parking, and construction.

Results have two supported depths. The listings depth contains search-level fields such as address, price, beds, baths, status, lot area, living area, and listing-type flags. The enriched depth is the default and adds detail-page and historical fields. The repository also includes sample CSV and JSONL data, Python and Node examples, and a workflow notebook, but those files are documentation and examples rather than additional MCP tools.

How it works

Requests are sent to the Apify actor through either Apify's hosted MCP endpoint at https://mcp.apify.com or the local @apify/actors-mcp-server package. The repository describes the MCP interface as a one-tool server. An agent can use catalog mode to request a cached, already-enriched snapshot, such as a Chicago catalog, without supplying a bounding box or waiting for a live scrape.

For current collection jobs, the actor supports custom_search and recent_activity modes in addition to catalog. A live workflow follows a fire, poll, and fetch pattern. Each run writes status information to its key-value store, including a plain-language explanation for zero-row or partial results.

The actor can also write a DEDUP_UPDATE record containing previously known and newly delivered Zillow property IDs and MLS IDs. Supplying those arrays as dedupZpids and dedupMlsIds on a later order helps avoid receiving and paying for duplicate rows.

Setup and configuration

Use an Apify account and connect the MCP client to Apify's hosted MCP service, or run the local actors MCP server. The repository specifically identifies Claude Desktop and Cursor as supported clients. The actor can also be run directly from the Apify Console or from Python or Node with apify-client.

A minimal catalog request uses a mode and metro, for example:

{"mode":"catalog","metro":"Chicago"}

For live orders, configure the actor input according to the actor's input schema, including the selected mode, result depth, and optional deduplication arrays.

Tools and capabilities

The zillow-leads-property-data MCP server supports access to the actor's listing-collection workflow, including:

• Requesting cached catalog snapshots.
• Starting live Zillow collection jobs.
• Retrieving bare or enriched listing rows.
• Accessing agent, broker, phone, email, brokerage, and MLS attribution fields when present.
• Reading price-history and 20-year-plus tax-history timelines.
• Filtering or analyzing foreclosure, distress, school, HOA, and resoFacts fields in returned data.
• Passing prior Zillow and MLS IDs to reduce duplicate deliveries.

Limitations and notes

The repository is documentation and sample code; it does not run the actor itself. Data collection and billing occur through Apify and the linked actor. Catalog mode uses a cached snapshot, so it is not the same as requesting a fresh search.

The zillow-leads-property-data MCP server does not document a broader set of custom tools beyond the one-tool actor interface. Field availability depends on the selected depth and the data returned by Zillow. For a zero-row or partial run, inspect the run's STATUS record before opening an issue. The material does not state a software license for the repository.