MCP Server
The UK Government Data MCP Server lets any AI agent query 28 official UK data sources through the Model Context Protocol. Works with Claude, GPT, and any MCP-compatible application.
Installation
Setup with Claude Desktop
Add to your claude_desktop_config.json:
Restart Claude Desktop. The tools will appear automatically.
Setup with Claude Code
Or add to ~/.claude/settings.json for persistent access:
Available Tools
lookup_postcode
Returns all 28 data sources for a UK postcode in one call.
Example prompt: “Tell me everything about SW11 5QD”
lookup_place
Resolves a place name, neighbourhood, postcode district, or landmark to a ward centroid postcode. Use this when the user provides a place name rather than a full postcode.
Example prompt: “What’s Clapham like?”
lookup_coordinates
Reverse geocodes latitude/longitude to the nearest postcode and returns resolved location info.
Example prompt: “What’s at coordinates 51.46, -0.175?”
get_crime
Crime statistics from the Home Office / data.police.uk. Returns 14 crime categories at street level with investigation outcomes.
Example prompt: “What’s the crime rate near E1 6AN?”
get_cleanliness
Resident-reported street conditions from FixMyStreet. 7 million reports covering fly-tipping, potholes, graffiti, broken street lights — hyper-local signal from actual residents.
Example prompt: “How clean is the street scene near SE15 4ST?”
get_house_prices
HM Land Registry Price Paid transactions, EPC energy ratings, and council tax band estimates.
Example prompt: “What have houses sold for near N1 9GU?”
get_demographics
Census 2021 demographics, population estimates, IMD deprivation index, household income, and labour market data (ONS / Nomis).
Example prompt: “What are the demographics of SE1 9SG?”
get_schools
Nearby schools with type, phase, age range, and Ofsted rating from GIAS.
Example prompt: “What schools are near W4 5TF?”
get_transport
TfL nearest tube/rail stations, bus stops, and commute times. Also EV charging (Open Charge Map) and road traffic (DfT).
Example prompt: “How well connected is SW19 5AE?”
get_environment
Air quality (LAQN / DEFRA), flood risk (Environment Agency), carbon intensity (National Grid ESO), and nearby green spaces (OSM).
Example prompt: “What’s the environmental risk for SE10 0DX?”
get_planning
Planning applications from the Planning Portal, conservation areas, listed buildings, and Article 4 directions.
Example prompt: “Any planning applications near W1D 3QF?”
get_local_services
Food hygiene ratings (FSA), GP practices (NHS ODS), supermarkets, and parliamentary constituency with current MP.
Example prompt: “What services are near EC2R 8AH?”
get_care
CQC-registered care homes, nursing homes, and domiciliary care providers with inspection ratings (Outstanding / Good / Requires Improvement / Inadequate) and bed counts.
Example prompt: “What care facilities are near SW1A 1AA?”
get_public_health
Public health indicators from OHID Fingertips (Dept. of Health and Social Care). Returns life expectancy, healthy life expectancy, obesity, smoking, depression, cardiovascular and cancer mortality, alcohol admissions, diabetes, physical inactivity, and more — all at local authority level.
Example prompt: “What are the public health indicators for residents of E14?”
get_companies
Active registered companies near a UK location from Companies House. Returns company names, types, SIC industry codes, and incorporation dates. Useful for commercial activity analysis, insurance underwriting, and investment DD.
Example prompt: “What businesses are registered near EC2V 8RF?”
lookup_company
Full Companies House profile for a specific company number. Returns company status, registered address, active officers (directors/secretaries), persons with significant control (PSCs), and outstanding charges/mortgages.
Example prompt: “Look up company number 09999999 — who are the directors and are there any outstanding charges?”
How it works
The MCP Server is a thin wrapper around the Speak to Scout REST API. Each tool call queries the unified endpoint and returns the relevant subset of data for the agent to interpret.
Source code
github.com/benanthoney-97/uk-gov-data-mcp
Licence
MIT
