mcpField
U
uk-parliament-erskinemay

uk-parliament-erskinemay

Local
github.com/DarkhorseOneLegal & Government1Updated 33w ago

About uk-parliament-erskinemay

An MCP server for querying UK Parliament Erskine May procedural content. It supports lookups and searches across chapters, sections, paragraphs, and index terms for parliamentary and legal research.

Key features

  1. 01Retrieve chapters, parts, and sections
  2. 02Search paragraphs and sections by keyword
  3. 03Browse and search index terms
  4. 04Look up section overviews by reference
  5. 05Paginate search results

Use cases

  1. 01Answer procedural questions with references
  2. 02Research parliamentary rules and sections
  3. 03Enrich civic datasets with Erskine May references
  4. 04Index procedural content for retrieval

MCP support

Exposes tools for querying the UK Parliament Erskine May API to MCP clients, including content retrieval, index browsing, and keyword search. Supports stdio transport.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"nick-ma-uk-parliament-erskinemay": {
"command": "npx",
"args": [
"-y",
"@darkhorseone/mcp-server-uk-parliament-erskinemay"
]
}
}
}
Run commandnpx -y @darkhorseone/mcp-server-uk-parliament-erskinemay

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@darkhorseone/mcp-server-uk-parliament-erskinemay1.0.1

Tools

11

Saved tool list · updated 2w ago

ParameterTypeRequiredDescription
chapterNumbernumberYesChapter overview with the chapter number specified

Details

From the project README

MCP Compatible AI Agent Ready

UK Parliament Erskine May MCP Server

Summary

This MCP server provides structured access to UK Parliament Erskine May content using the Model Context Protocol. It enables AI agents to query chapters, parts, sections, and index terms through API-backed tools for reliable data access. Developers can connect the MCP server via stdio transport to retrieve authoritative procedural guidance and references. The server exposes tools that map directly to the Erskine May API for search and lookup workflows.

Features

  • Retrieve chapters, parts, sections, and section overviews by identifier.
  • Browse and search index terms and paragraphs by keyword.
  • Support structured lookups with consistent response envelopes.
  • Preserve upstream JSON payloads for downstream tools and AI agents.

Available Tools

get_chapter_by_chapter_number

Returns a single chapter overview by chapter number.

Parameters:

  • chapterNumber (number, path) – Chapter number.
browse_index_terms

Returns a list of index terms by start letter.

Parameters:

  • startLetter (string, query)
  • skip (number, query)
  • take (number, query)
get_index_term_by_index_term_id

Returns an index term by id.

Parameters:

  • indexTermId (number, path) – Index term id.
list_parts

Returns a list of all parts.

get_part_by_part_number

Returns a part by part number.

Parameters:

  • partNumber (number, path) – Part number.
search_index_term_results_by_search_term

Returns a list of index terms which contain the search term.

Parameters:

  • searchTerm (string, path)
  • skip (number, query)
  • take (number, query)
get_paragraph_by_reference

Returns a section overview by reference.

Parameters:

  • reference (string, path)
search_paragraph_results_by_search_term

Returns a list of paragraphs which contain the search term.

Parameters:

  • searchTerm (string, path)
  • skip (number, query)
  • take (number, query)
search_section_results_by_search_term

Returns a list of sections which contain the search term.

Parameters:

  • searchTerm (string, path)
  • skip (number, query)
  • take (number, query)
get_section_by_section_id

Returns a section by section id.

Parameters:

  • sectionId (number, path)
get_section_overview_by_section_id_and_step

Returns a section overview by section id and step.

Parameters:

  • sectionId (number, path)
  • step (number, path)

Example Output

{
  "status": 200,
  "data": {
    "sectionId": 12,
    "title": "Order in debate",
    "content": "..."
  },
  "upstream_path": "/api/Section/12",
  "retrieved_at": "2026-03-11T10:00:00.000Z"
}

Quick Start

Run the MCP server using npx:

npx -y -p @darkhorseone/mcp-server-uk-parliament-erskinemay mcp-server-uk-parliament-erskinemay

MCP Configuration

Example mcpServers configuration for stdio transport:

{
  "mcpServers": {
    "uk-parliament-erskinemay": {
      "command": "npx",
      "args": ["-y", "-p", "@darkhorseone/mcp-server-uk-parliament-erskinemay", "mcp-server-uk-parliament-erskinemay"],
      "env": {
        "UPSTREAM_BASE_URL": "https://erskinemay-api.parliament.uk",
        "REQUEST_TIMEOUT_MS": "10000"
      }
    }
  }
}

Example Usage

  • Fetch a chapter by chapter number.
  • Search sections for a keyword and paginate results.
  • Retrieve a section overview by id and step.
  • Browse index terms starting with a letter.

Use Cases

  • AI agents answering procedural questions with authoritative references.
  • Legal or parliamentary research tools mapping sections and paragraphs.
  • Civic tech workflows enriching datasets with Erskine May references.
  • Automation systems that index procedural content for retrieval.

Data Source

UK Parliament Erskine May API https://erskinemay-api.parliament.uk

Installation

Run directly using npx:

npx -y -p @darkhorseone/mcp-server-uk-parliament-erskinemay mcp-server-uk-parliament-erskinemay

Or install via pnpm from the repository root:

pnpm install

Run (HTTP proxy server)

Build first, then start:

pnpm --filter @darkhorseone/mcp-server-uk-parliament-erskinemay run build
pnpm --filter @darkhorseone/mcp-server-uk-parliament-erskinemay run start:http

Set a custom HTTP port with the UKP_HTTP_PORT environment variable (default: 3000):

UKP_HTTP_PORT=3000 pnpm --filter @darkhorseone/mcp-server-uk-parliament-erskinemay run start:http

Published package command:

npx -y -p @darkhorseone/mcp-server-uk-parliament-erskinemay -- mcp-server-uk-parliament-erskinemay-http

If your shell cannot resolve the bin ("command not found"), run the HTTP entrypoint directly:

npx -y -p @darkhorseone/mcp-server-uk-parliament-erskinemay -- node ./dist/http.js

Health check:

curl http://127.0.0.1:3000/healthz

Proxy pattern:

GET /proxy/<upstream-path>?<query>

Example:

curl "http://127.0.0.1:3000/proxy/api/Part"

License

MIT License

MCP Metadata

Protocol: Model Context Protocol Transport: stdio Tools: get_chapter_by_chapter_number, browse_index_terms, get_index_term_by_index_term_id, list_parts, get_part_by_part_number, search_index_term_results_by_search_term, get_paragraph_by_reference, search_paragraph_results_by_search_term, search_section_results_by_search_term, get_section_by_section_id, get_section_overview_by_section_id_and_step

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.