Uk Caselaw MCP Server

UK Case Law — Find Case Law

Judgments and tribunal decisions of the UK courts, from The National Archives' Find Case Law service: the Supreme Court, Privy Council, Court of Appeal, High Court, Court of Protection, Family Court, the Upper and First-tier Tribunals and the Employment Appeal Tribunal.

Part of Pipeworx — an MCP gateway connecting AI agents to 1866+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Search by subject, party or judge, then read the full text of any judgment from its neutral citation — so a quoted passage or a cited authority can be checked against what the court actually wrote.

This is UK law only. US case law is a separate corpus served by court-listener (search_case_law, find_case, get_opinion); an American authority returned for a UK question is a wrong answer, not a near miss.

This is not all UK law

Find Case Law is a real but partial corpus. A 0-row result means not found here, never no such case:

• No Crown Court, County Court or Magistrates' Court judgments. Most of what those courts decide is given orally and never transcribed, so Find Case Law structurally cannot hold it.
• No dockets, pleadings or filings. This is judgments only — there is no UK equivalent of PACER here.
• Scotland and Northern Ireland's own court systems are not covered. Only England & Wales courts and the UK-wide apex courts/tribunals (Supreme Court, Privy Council, Employment Appeal Tribunal, Upper Tribunal) are indexed. A case that also went through the Court of Session or a Scottish/NI appeal may still appear here only for its UK-level leg.
• The Employment Appeal Tribunal only — not the first-tier Employment Tribunal. eat covers appeals; ET claims before they reach appeal are a different, uncovered corpus (see the uk-tribunals pack/task).
• No citator/treatment data. A judgment comes back as published, with no signal on whether it was later overturned, doubted or followed.
• Coverage is mostly 2001 onward, comprehensively from 2003 for some courts and much later for others (see Caveats below).

Auth

None. No key, no signup. Requests carry a browser User-Agent because the service rejects a bare fetch signature.

Tools

The usual path is search_uk_caselaw → take the neutral_citation or identifier → get_uk_judgment.

Neutral citations

A neutral citation maps onto a Find Case Law path by rule — [2026] EWHC 1996 (Comm) is ewhc/comm/2026/1996, [2023] UKSC 42 is uksc/2023/42. The rule is a good first guess, so get_uk_judgment tries it and falls back to searching the citation text when the derived path 404s, reporting which route resolved it in resolved_by. Either form is accepted as input.

Data sources

• Search: https://caselaw.nationalarchives.gov.uk/atom.xml (Atom feed)
• Judgment text: https://caselaw.nationalarchives.gov.uk/<identifier>/data.xml (Akoma Ntoso XML)
• Service: https://caselaw.nationalarchives.gov.uk/

Judgments are Crown copyright, published under the Open Justice Licence.

Licence note. The Open Justice Licence covers per-query pass-through — a caller searches, we relay the matching judgments back — but it explicitly excludes "computational analysis": bulk programmatic extraction or mirroring/indexing the corpus requires a separate licence Pipeworx does not hold. This pack is not, and must not become, a mirror of Find Case Law.

Caveats worth passing to a user

• Coverage starts recently. Find Case Law holds judgments the courts published to it — comprehensively from 2003 for some courts and much later for others. A missing older authority means not held here, not no such case. For historic authority a law-report subscription is still the answer.
• An unrecognised court code is rejected before any request goes upstream, as an invalid_arguments error naming the bad code and the full valid list (call list_uk_courts to get it fresh). Codes look like uksc, ewca/civ, ewhc/comm and cannot be guessed from a court's name — and the Employment Appeal Tribunal's code is eat, not the "ukeat" its name would suggest (fixed 2026-09-18; every code in list_uk_courts is now one that has been probed live against the atom feed and confirmed to return entries).
• Search is over the judgment text, so a phrase the court would actually write finds more than a legal concept it never names.
• Judgments are long — 140k–450k characters is ordinary. get_uk_judgment bounds the response with max_chars and sets truncated so a cut-off text is never mistaken for the whole judgment.
• Only newest, oldest and relevance sort. Upstream answers an unrecognised sort with an empty feed, which would read as "no such cases", so the pack rejects one instead of passing it through.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "uk-caselaw": {
      "url": "https://gateway.pipeworx.io/uk-caselaw/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/uk-caselaw/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1866+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/search_uk_caselaw \
  -H 'Content-Type: application/json' \
  -d '{"query":"unjust enrichment","court":"uksc","limit":3}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/search_uk_caselaw. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Run it from npm

@pipeworx/mcp-uk-caselaw is a local stdio bridge to the Pipeworx gateway: your MCP client starts it with npx, and it forwards every call to the same https://gateway.pipeworx.io/uk-caselaw/mcp endpoint shown above. It needs network access; there is no offline mode.

{
  "mcpServers": {
    "uk-caselaw": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-uk-caselaw"]
    }
  }
}

Without a key it calls anonymously, capped at 50 calls a day. A free account raises that to 200 a day: get a key at https://pipeworx.io/signup and add "env": { "PIPEWORX_API_KEY": "<your key>" } to the server entry. The key is sent as a Bearer token, the same way a hosted connection sends it, so calls are metered and billed exactly as they are there.

It lists only this pack's tools by default. Set PIPEWORX_META_TOOLS=1 to also list ask_pipeworx and the other gateway-wide tools.

Behind an HTTPS proxy: Node's built-in fetch does not read HTTPS_PROXY by default. Newer Node releases honour it when NODE_USE_ENV_PROXY=1 is set; check your Node version's documentation.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Uk Caselaw data" })

The gateway picks the right tool and fills the arguments automatically.

More

• Docs and guides
• pipeworx.io

License

MIT