Ordnance Survey - MCP Server
VERSION: 0.1.1
A Python-based MCP server that provides access to the Ordnance Survey APIs, supporting both STDIO and HTTP (streamable) modes.
Overview
This service creates a bridge between MCP clients and OS DataHub APIs - making it easy to query national geographic data through a standardised protocol.
It can run in two modes:
- STDIO mode: Ideal for Claude Desktop and local tool integration such as Cursor
- HTTP (streamable) mode: Perfect for web clients/applications
Project Structure
- api_service - Asynchronous HTTP client implementation for the OS APIs
- Handles authentication, request formatting, and response processing
- Manages rate limiting and error handling
- Provides a clean interface to the external API endpoints
- mcp_service - Exposes OS APIs functionality as MCP tools
- Converts API responses to MCP-compatible formats
- Implements business logic for feature operations
- Provides a standardised interface
- middleware - Middleware for the MCP server
stdio_middleware.py
: Handles authentication for STDIO transporthttp_middleware.py
: Handles authentication for HTTP transport
- prompt_templates - Prompt templates for common operations
- Provides pre-configured prompt templates to help you get started with some common operations
- config_docs - Documentation for OS APIs
- Provides documentation for the OS APIs
- utils - Utility functions for the MCP server
- Provides utility functions for the MCP server such as logging
Features
- Collection management (listing and querying collections)
- Feature search with spatial and attribute filters
- Individual feature retrieval by ID
- Linked identifier operations
- Bulk feature operations
Requirements
- Python 3.11+
- OS API Key (set as environment variable
OS_API_KEY
) - You will need to register for an OS Data Hub account to get an API key
Running the Server
1. STDIO Mode (for Claude Desktop)
This is the default mode, ideal for integration with Claude Desktop or other MCP hosts that use STDIO.
- Configure your MCP host (e.g., in Claude Desktop's configuration):
- Start the server manually (for testing):
2. HTTP (Streamable) Mode
This mode is ideal for web clients or when you need to stream large datasets.
You will need to set the OS_API_KEY
and BEARER_TOKEN
environment variables.
Each request to the MCP server will need to be authenticated with a bearer token - it's currently set to dev-token
in the client test script.
- Start the server:
- Test using the provided client script:
Note: The client test script (
client_test.py
) is a great way to verify your server setup and see example code for programmatic interaction with the API. It uses themcp.client.streamable_http
library to demonstrate proper connection handling and tool calling.
The client script demonstrates:
- Connecting to the MCP server
- Initialising a session
- Listing available tools
- Making test calls (e.g.,
hello_world
tool)
Available Tools
All tools are available in both STDIO and HTTP modes:
hello_world
- Test connectivitycheck_api_key
- Verify API key configurationlist_collections
- List available feature collectionsget_collection_info
- Get details about a specific collectionget_collection_queryables
- Get filterable properties for a collectionsearch_features
- Search features by various criteriaget_feature
- Retrieve a specific feature by IDget_linked_identifiers
- Find related identifiersget_bulk_features
- Retrieve multiple features in a single callget_bulk_linked_features
- Get linked features in bulkget_prompt_templates
- Get standard prompt templates for common operationssearch_by_uprn
- Search for addresses by UPRNsearch_by_post_code
- Search for addresses by POSTCODEget_map_tile
- Get a map tile in EPSG:27700 projection - THIS DOES NOT WORK - NEED TO FIX
Using Prompt Templates
This service provides pre-configured prompt templates to help you get started.
To access these templates ask Claude "show me available prompt templates"
Current prompt templates are:
connected_usrns
- Find all USRNs that are directly connected to a given USRN
Contributing
Contributions are welcome! Please open an issue or submit a pull request.
License
This project is licensed under the MIT License.
This project does not have the endorsement of Ordnance Survey.
This server cannot be installed
remote-capable server
The server can be hosted and run remotely because it primarily relies on remote services or has no dependency on the local environment.
A Python-based MCP server that provides access to Ordnance Survey APIs, allowing querying of geographic data through a standardized protocol with features like collection management, feature search, and spatial filtering.
Related MCP Servers
- AsecurityAlicenseAqualityThe MCP server provides an interface to the Datadog API, enabling seamless management of incidents, monitoring, logs, dashboards, metrics, traces, and hosts. Its extensible design allows easy integration of additional Datadog APIs for future expansions.Last updated -1432236TypeScriptApache 2.0
- AsecurityFlicenseAqualityAn MCP server implementation that enables interaction with the Unstructured API, providing tools to list, create, update, and manage sources, destinations, and workflows.Last updated -3928
- -securityFlicense-qualityAn MCP server that provides access to Paylocity API endpoints, allowing retrieval of employee data, earnings, company codes, local taxes, and pay statements.Last updated -Python
- -securityFlicense-qualityA Python-based server that helps users easily install and configure other MCP servers across different platforms.Last updated -2Python