How AI agents track UK government announcements and court judgments
An agent tracks UK government activity through two separate feeds: GOV.UK for departmental announcements, policy papers and guidance, and Find Case Law for court judgments from England and Wales plus UK-wide Supreme Court cases. They are different publishers with different coverage, so query them separately.
/mcp. Nothing here needs an API key or a signup.
Agent tooling skews heavily toward US sources, which leaves UK coverage underserved relative to how much of it is published openly. Both of these publishers are unusually machine-friendly by government standards.
What each publisher actually covers
| GOV.UK | Find Case Law | |
|---|---|---|
| Publisher | UK government departments | The National Archives |
| Content | News, policy papers, guidance, consultations | Court judgments and tribunal decisions |
| Jurisdiction | UK government, so mostly UK-wide | England and Wales, plus UK-wide apex courts |
| Rhythm | Working days, steady | Working days, bursty by court |
| Wire | /uk/latest at $0.005 | /ukcourt/judgments at $0.01 |
The split is not cosmetic. A question about what the government has decided goes to the first; a question about what a court has held goes to the second. An agent that queries only GOV.UK for a legal question will return a press release about a case rather than the judgment.
Where the coverage genuinely stops
This is the part most descriptions skip, and an agent that does not know it will report absence as fact.
Find Case Law provides free access to judgments from courts in England and Wales, plus UK-wide cases from the Supreme Court and the Privy Council. Its own description of scope is candid: coverage for most courts starts in the early 2000s, when digital records became standard, and the Supreme Court itself was only established in 2009. It holds the vast majority of what courts have made publicly available in digital form.
Two limits follow, and both are worth an agent stating rather than papering over:
- Older case law is thin. A search that returns nothing from 1985 has found
a gap in digital coverage, not an absence of law.
- Not every decision is written down at all. Many lower court hearings and
some tribunal decisions are given verbally and never transcribed, so they cannot appear in any feed.
Scotland and Northern Ireland have their own court systems and are not covered by this collection. An agent answering a UK legal question should say which jurisdiction it actually searched.
Querying them together without merging them
The useful pattern is two calls and one labelled answer, rather than one merged result set. Government position and judicial holding are different kinds of claim, and an agent that flattens them loses the distinction a reader needs.
# UK government announcements, policy and guidance curl -i "https://thebotwire.com/uk/latest?q=online+safety&src=govuk&since=7d" # court judgments, England and Wales plus UK apex courts curl -i "https://thebotwire.com/ukcourt/judgments?q=data+protection&since=30d&limit=25" # free, no wallet: connect the MCP server claude mcp add --transport http botwire https://thebotwire.com/mcp
Both wires are sparse sources with a 7 day default window, so widen before concluding a quiet week means nothing happened. For the neighbouring European regulatory picture see how agents follow European Commission decisions, and for the US equivalent of the judgments wire see how agents read federal court opinions. Full parameters for every wire are at /docs.
FAQ
Does GOV.UK have its own API an agent could call directly?
Yes. GOV.UK publishes a content API, and for a developer who knows in advance that they need it, calling it directly is the reasonable choice. A wire earns its cost when the need appears at runtime, or when one query should span UK government output alongside courts and other jurisdictions.
Are Scottish and Northern Irish cases included?
No. Find Case Law covers England and Wales, plus UK-wide cases from the Supreme Court and Privy Council. Scotland and Northern Ireland maintain separate court systems with their own publication arrangements, and an agent should scope its answer accordingly rather than implying UK-wide coverage.
How far back does the judgment archive go?
For most courts, the early 2000s, when digital records became standard practice. The Supreme Court's own record begins with its creation in 2009. Anything earlier needs a commercial legal database, and an empty result for a historic case is a coverage gap rather than evidence about the law.
Why does the court wire cost more than the GOV.UK wire?
The same reason the US court and filings wires do: the documents are longer, the parsing is heavier, and the answers they support are worth more per call. Both prices are stated in the 402 response before any payment happens, and are listed at /health.
Sources
Coverage scope, the early 2000s digital start point, the 2009 establishment of the Supreme Court, and the note that untranscribed decisions cannot be published are stated by The National Archives at Find Case Law. The GOV.UK content API is documented at content-api.publishing.service.gov.uk. Poll cadence and per-call pricing for both wires are published at thebotwire.com/health.
Related: How do AI agents follow European Commission decisions? · How do AI agents read new federal court opinions?