`list_jurisdictions`

Per-country schema reference and cross-country tool-support matrix for every national company registry OpenRegistry covers.

Reference-data tool. Pass a jurisdiction code to get that country's full schema (registry name, ID format, supported tools, native status values, known quirks, jurisdiction_data field catalogue). Pass a tool name to get the cross-country support matrix for that specific tool.

Parameters

Name	Type	Required	Description
`jurisdiction`	`string`	no	ISO 3166-1 alpha-2 country code. Mutually exclusive with supports_tool.
`supports_tool`	`string`	no	Tool name. Returns {supported_count, supported_in[], not_supported_in[]}. Mutually exclusive with jurisdiction.

Example - Get full GB schema

curl -sL https://openregistry.sophymarine.com/mcp \
  -H 'content-type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"list_jurisdictions","arguments":{"jurisdiction":"GB"}}}'

Supported jurisdictions

This tool currently has no live registry bindings.

Related tools

Frequently asked questions

When should I call list_jurisdictions vs hardcode a country list?

Call it once per AI conversation for any tool you're about to use against an unknown country. Result is stable in-conversation, free, no rate cost. Hardcoding ISO codes is fine for known-target queries but misses dynamically added jurisdictions and silently fails when a country's tool support changes.

What does the per-jurisdiction response include?

supported_tools (which of the 26 tools that country exposes), company_id_format (with examples), registry_name + registry_url (the official upstream), data_license, native status values + jurisdiction_data field catalogue, real_time flag, caching_notes, and known_quirks (registry-specific gotchas the AI should account for before constructing queries).

What's supports_tool for?

Cross-country matrix mode. supports_tool: 'get_shareholders' returns {supported_count: 13, supported_in: ['CH','DE',...], not_supported_in: ['GB','IE',...]}. Faster than calling per-country when you want to know 'which countries can I run THIS tool on'.

Are both jurisdiction and supports_tool allowed at once?

No — mutually exclusive, server rejects. Either you want one country's schema, or one tool's coverage matrix. To get both, two calls.

Does the response change between calls?

Stable within a release cycle. New jurisdictions / tool wirings are deployed in batches with announcements in the changelog. In-conversation cache the response — it won't change mid-session.

Why is country X missing?

Either not yet wired (most often: paid-API gating, e.g. DE for accounts, several EU UBO registers) or temporarily offline (DE Handelsregister scrapes flaky as of 2026-05). Live count returned in result.count; compare against the marketing page's '27 jurisdictions' for the gap.

What is 'company_id_format'?

A free-text spec + 3–6 example IDs that pass validation. GB: '8-digit Companies House number with optional leading SC/NI/FC/OC prefix', examples ['09446231','SC123456']. Tells the AI how to construct a query before search_companies.

What are 'known_quirks'?

Per-country gotchas. E.g. GB: 'company numbers can begin with SC (Scotland) / NI (Northern Ireland) / OC (LLP) / FC (overseas branch)'. KR: 'corp_code is an 8-digit DART internal ID, distinct from the 13-digit business registration number'. Read before constructing a query against an unfamiliar country.

How big is the response?

Per-country mode ~3 KB. supports_tool matrix ~1 KB. Top-level list (no params): ~5 KB summary of all jurisdictions. Cheap to call freely.

Is this the same as the /llms.txt manifest?

Overlapping but not identical. /llms.txt is the human-readable static manifest for AI crawlers (llmstxt.org spec). list_jurisdictions is the runtime tool — returns the live capability matrix at the moment of call, faster to parse for an AI agent than markdown.