US & UK Public Records MCP for AI Agents

An MCP server that gives Claude, ChatGPT, Gemini, Cursor and any other MCP client nine tools on official public records, straight from government open data:

Tool	Official source	What the agent gets
find_contractor_licenses	Washington State L&I, New York State DOL	Contractors newly licensed in Washington or registered for public works in New York: business, license type and specialties, phone, city, principals, license number, status and a link to verify it.
find_healthcare_providers	CMS NPI registry (NPPES)	Doctors, dentists, therapists, clinics and pharmacies by specialty and place, or one NPI number to verify: specialty, credential, practice phone and fax, address, authorized official.
find_trucking_companies	FMCSA census	Trucking companies, new ones first, or looked up by USDOT number: legal name, phone, email, officers, fleet, drivers, operation, hazmat and cargo.
search_federal_contracts	USAspending.gov	New federal contract awards: winning company, amount, agency, NAICS, description and place of performance.
restaurant_inspections	Health departments of New York City, Chicago, King County WA, Montgomery County MD, Cincinnati and Austin	Inspection results, grades and scores, violations, address and phone, and new restaurants inspected before they open.
search_building_permits	Building departments of New York City, Chicago, San Francisco, Seattle, Cincinnati, Austin and Baton Rouge	Permits by type of work (solar, roofing, ADU, pool, remodel...), project value, address, contractor with phone, owner and status.
find_new_us_businesses	Secretaries of State of NY, CO, CT, OR and PA; Los Angeles, San Francisco and Chicago	Newly registered LLCs, corporations and DBAs with address, owners or officers, registered agent, and email and NAICS where published.
uk_sold_house_prices	HM Land Registry Price Paid Data	Homes sold in England and Wales by postcode, town, council or county: price, date, address, type, tenure and new build.
uk_food_hygiene_ratings	Food Standards Agency	UK food businesses with their official hygiene rating and its date, and new openings waiting for their first inspection.

Ask your agent things like "Is there a dentist practice in Austin ZIP 787 that opened this year? Give me its phone", "Check the carrier with USDOT 3456789", "Who won cybersecurity contracts from the federal government this week?", "Did BEBECITAS in Brooklyn pass its health inspection?", "Which solar permits were issued in Austin in the last two weeks, and by which installers?" or "What did flats in SW11 sell for this year?": it picks the tool and gets clean JSON back, newest records first.

Connect your agent

Easiest, with your Apify account (sign-in in the browser, no token to copy): add this remote MCP server URL to your client:

https://mcp.apify.com?tools=digital_influx/public-records-mcp

• Claude (claude.ai, Claude Desktop): add a custom connector with that URL.
• ChatGPT: add it as a connector (MCP server URL) in developer mode.
• Cursor, VS Code, Windsurf, Gemini CLI and other clients: add a remote MCP server. With a token instead of the browser sign-in:

{
  "mcpServers": {
    "public-records": {
      "url": "https://mcp.apify.com?tools=digital_influx/public-records-mcp",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

(Gemini CLI writes httpUrl instead of url in ~/.gemini/settings.json.)

Directly, without the Apify MCP server: this Actor runs in Standby mode as a Streamable HTTP MCP server. Take its Standby URL from the Actor's API tab in Apify Console, add /mcp, and send your Apify token as Authorization: Bearer YOUR_APIFY_TOKEN.

A record of every call: each tool call your agent makes is also saved as a row of the Standby run's dataset (tool, arguments, a one-line summary and the result), so you can check later what it asked and got. Without an agent: a standard run calls one tool from the input (tool and arguments) and saves the result to the dataset, so the tools also work from the API, schedules and integrations (Make, n8n, Zapier).

Arguments

Every tool takes the filters of the Actor it runs, with the same names, choices and limits, plus limit (records to return, 20 by default, at most 100). For example:

{ "taxonomies": ["Dentist"], "zipCodes": ["787"], "enumeratedWithinDays": 365, "limit": 20 }

{ "keywords": ["cybersecurity"], "postedWithinDays": 7, "minAmount": 100000 }

{ "postcodes": ["SW11"], "propertyTypes": ["flat"], "soldWithinMonths": 6 }

Your MCP client shows the full list of each tool's arguments with their descriptions. For uk_sold_house_prices, limit is shared among the areas you give.

How much it costs

This server charges nothing itself. Each tool runs a public-records Actor of the same author in your account, and that run bills what it returns, at the same price as running it directly:

Tool	Actor	Price
find_contractor_licenses	US Contractor Licenses	USD 0.004 per license
find_healthcare_providers	US Healthcare Providers (NPI)	USD 0.003 per provider
find_trucking_companies	US Trucking Companies (FMCSA)	USD 0.004 per carrier
search_federal_contracts	US Federal Contracts	USD 0.005 per contract
restaurant_inspections	US Restaurant Inspections	USD 0.005 per inspection
search_building_permits	US Building Permits	USD 0.003 per permit
find_new_us_businesses	New Business Leads (US)	USD 0.004 per business
uk_sold_house_prices	UK Sold House Prices	USD 0.0015 per sale
uk_food_hygiene_ratings	UK Food Hygiene Ratings	USD 0.003 per business

A search with no results costs nothing. Runs in Standby use 256 MB, and each call takes 10 to 60 seconds while the Actor runs.

Good to know

• Official open data only, read through the public APIs the governments publish for it (Socrata portals, NPPES, USAspending, HM Land Registry, the FSA API), with an honest User-Agent and the request pace each portal asks for. No scraping of private sites, no proxies, no login.
• Records are what the agencies publish, with the delay each one has (most daily; Oregon and Pennsylvania business registries monthly; Land Registry sales reach the register weeks after completion). Phones and emails are the ones published in the registries: follow the TCPA, CAN-SPAM, the GDPR and PECR where they apply.
• The same Actors work on their own, for thousands of records in one batch, scheduled feeds of only the new records, and every field. Their pages list the sources and the attribution each one requires.
• More tools for agents by the same author: Company Research MCP (company contacts, the legal entity behind EU websites, tech stack, SEC filings) and Job Search MCP (jobs from 20,000+ company career sites).