Lawstronaut MCP

5.1 list_jurisdictions

Purpose. Enumerate every ISO code the MCP recognises. Always call this first when an LLM doesn't already know the valid ISO for the user's target country/state.

Parameters. none.

Live example (trimmed).

{
  "data": [
    { "name": "Belgium",        "iso": "BE",     "type": "country" },
    { "name": "European Union", "iso": "EU",     "type": "country" },
    { "name": "Ireland",        "iso": "IE",     "type": "country" },
    { "name": "United States",  "iso": "US",     "type": "country" },
    { "name": "California",     "iso": "US_CA",  "type": "state"   },
    { "name": "Bavaria",        "iso": "DE_BY",  "type": "state"   },
    { "name": "Brussels",       "iso": "BE_BR",  "type": "state"   }
    // ... 132 entries total
  ]
}

Notes.

• ISO uses underscore-prefixed sub-state codes (US_CA, not US-CA).
• "International Standards" is listed under ISO "ZZ" — convenient for ISO/IEC standards content with no national jurisdiction.
• The European Union is treated as a country-level jurisdiction (iso: "EU").

5.2 list_portals

Purpose. Discover the official portals (data sources) Lawstronaut has crawled for a given jurisdiction. Only portals that have actually produced crawled data are returned.

Parameters.

Live example (trimmed).

[
  {
    "name": "Belgium Etaamb Openjustice (French)",
    "name_en": "Etaamb Openjustice",
    "url": "etaamb.openjustice.be/fr",
    "language": "French",
    "jurisdiction": { "country": "Belgium", "state": "" },
    "portal_tags": ["General Legislation", "legal professionals", ...],
    "total_links": 346363
  },
  {
    "name": "JUPORTAL (Openbare databank voor Belgische rechtspraak)",
    "name_en": "JUPORTAL – Public Database of Belgian Case Law",
    "url": "juportal.be",
    "language": "Dutch",
    "portal_tags": ["General Case Law", "court and judiciary", ...],
    "total_links": 246433
  }
  // ... 45 Belgian portals total
]

Notes.

• total_links is the crawler-link count — a good proxy for how thoroughly the portal has been ingested.
• Many portals exist in multiple language variants ((Dutch), (French), (German)); choose lang deliberately when language matters.
• name is the canonical portal identifier used by list_documents(portal=...). Always use the exact name (case- and parenthesis-sensitive), not name_en.

5.3 list_domains

Purpose. Top-level legal-taxonomy buckets. Cached and tiny — there are only three.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 3, "limit": 10, "offset": 0 },
  "data": [
    { "domain_id": "A", "domain": "Private law"   },
    { "domain_id": "B", "domain": "Public Law"    },
    { "domain_id": "C", "domain": "Miscellaneous" }
  ]
}

5.4 list_subdomains

Purpose. Children of a domain.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 2, "limit": 10, "offset": 0 },
  "data": [
    { "domain_id": "A", "subdomain_id": "A.1", "subdomain": "Private International Law" },
    { "domain_id": "A", "subdomain_id": "A.2", "subdomain": "National law" }
  ]
}

5.5 list_categories

Purpose. Categories under a subdomain. subdomain_id is optional — calling without it lists all 81 categories across the corpus.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 81, "limit": 10, "offset": 0 },
  "data": [
    { "domain": "Private law", "subdomain": "Private International Law",
      "category_id": "A.1.1", "category_name": "General Private International Law" },
    { "domain": "Private law", "subdomain": "National law",
      "category_id": "A.2.5", "category_name": "Corporate law" }
    // ...
  ]
}

5.6 list_subcategories

Purpose. Subcategories under a category.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 21, "limit": 10, "offset": 0 },
  "data": [
    { "category_id": "A.2.5", "category_name": "Corporate law",
      "subcategory_id": "A.2.5.1",  "subcategory_name": "Company formation and incorporation" },
    { "category_id": "A.2.5", "category_name": "Corporate law",
      "subcategory_id": "A.2.5.16", "subcategory_name": "Restructuring, dissolution and bankruptcy" }
  ]
}

5.7 list_law_types

Purpose. Leaf-level law-type taxonomy.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 25, "limit": 10, "offset": 0 },
  "data": [
    { "law_type_id": 1507, "law_type": "Company Formation Policy and Framework Laws",
      "subcategory_id": "A.2.5.1", "category_id": "A.2.5", "category_name": "Corporate law" },
    { "law_type_id": 1508, "law_type": "Incorporation Procedures Laws", ... }
  ]
}

5.8 list_authority_types

Purpose. Enumerate the kinds of instrument (Act, Decree, Prospectus, Arrest, Reports, etc.) issued in a jurisdiction, optionally filtered by portal.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 368, "limit": 10, "offset": 0 },
  "data": [
    { "iso": "IE", "authority_type": "Circular Economy",     "portal_name": "Ireland Environmental Protection Agency" },
    { "iso": "IE", "authority_type": "Strategic statement",  "portal_name": "Competition and Consumer Protection Commission" }
  ]
}

Notes.

• Use this when you need to constrain a list_documents query to a specific instrument type (e.g. only Acts, only judgments).

5.9 list_issuing_authorities

Purpose. The named bodies (ministries, regulators, courts, agencies) that issued documents in a jurisdiction.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 272, "limit": 10, "offset": 0 },
  "data": [
    { "iso": "IE", "issuing_authority": "HSE board",
      "portal_name": "Health Service Executive" },
    { "iso": "IE",
      "issuing_authority": "Anti-Money Laundering Compliance Unit (AMLCU), Department of Justice (Ireland)",
      "portal_name": "Ireland Anti-Money Laundering Compliance Unit" }
  ]
}

Notes.

• Use this when the user names a regulator ("show me everything from the Belgian Data Protection Authority") and you need the exact string to feed into list_documents(issuing_authority=...).

5.10 list_documents

Purpose. The workhorse. Returns full document metadata filtered by any combination of 25+ optional criteria.

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 5999, "limit": 1, "offset": 0 },
  "data": [{
    "document_id": 20951014,
    "title": "Arrest nr. 24/2019 van 14 februari 2019",
    "jurisdiction": { "country": "Belgium", "state": "" },
    "publication_date": "2019-02-14T00:00:000Z",
    "issuing_authority": "Constitutional Court",
    "type_of_authority": "Arresten",
    "language": "Dutch",
    "legal_link": "https://nl.const-court.be/public/n/2019/2019-024n.pdf",
    "repealed": false,
    "status": "new", "version": 1,
    "source_identifier": "Rolnummer: 6776",
    "source_secondary_identifier": "ECLI:BE:GHCC:2019:ARR.024",
    "file_data": {
      "content_type": "application/pdf",
      "file_size": 116081,
      "timestamp": "2026-02-27T02:04:34.369535"
    }
  }]
}

Notes.

• Date timestamps come back as YYYY-MM-DDTHH:MM:SS00Z (note the trailing 0 before Z — a serialization quirk; parse defensively).
• ECLI identifiers come through source_secondary_identifier.
• file_data: {} (empty) means no binary; populated means a PDF is available via get_source_url.
• The same document_id can be fed to get_document_text, get_markdown, get_source_url, and get_document_with_version.

5.11 get_document_text

Purpose. Return the plain-text rendition of one or more documents. Accepts the same superset of filters as list_documents.

Parameters.

Live example (trimmed).

{
  "total_count": 1, "limit": 1, "offset": 0,
  "data": [{
    "document_id": "27732019",
    "full_text": "Number 7 of 2026 PROTECTION OF EMPLOYEES (EMPLOYERS' INSOLVENCY)
(AMENDMENT) ACT 2026 An Act to amend the Protection of Employees ...
[full text continues]"
  }]
}

Notes.

• document_id is a string in this response even though it is an integer in list_documents. Cast defensively.
• Plain text is unformatted (no markdown, no inline links). For richer rendering use get_markdown.
• Sweeping calls with broad filters can return large payloads — pass limit deliberately.

5.12 get_markdown

Purpose. Same content as get_document_text but as markdown — bold/italic preserved, inline cross-references rendered as [text](url).

Parameters.

Live example (trimmed).

*Number* 7 *of* 2026
**PROTECTION OF EMPLOYEES (EMPLOYERS' INSOLVENCY) (AMENDMENT) ACT 2026**
An Act to amend the
[Protection of Employees (Employers' Insolvency) Act 1984](https://www.irishstatutebook.ie/1984/en/act/pub/0021/index.html)
; to give further effect to Directive 2008/94/EC ...

Notes.

• Preferred over get_document_text whenever the downstream consumer can render markdown.
• Markdown is the canonical store; get_document_text is essentially get_markdown with formatting stripped.

5.13 get_document_with_version

Purpose. Retrieve a specific version of a specific document.

Parameters.

Live example (trimmed).

// Expected: single document object mirroring a list_documents row.
// Observed 2026-05-27: calls with version=1 against documents whose current version
// is 1 returned {} (empty). See Known Quirks — open question for the API team.

Notes.

• Working theory: the endpoint may only return historical versions, not the latest. Until confirmed, prefer list_documents(document_id=..., version=...).

5.14 get_source_url

Purpose. Get a time-limited presigned URL to the original source file (typically a PDF).

Parameters.

Live example (trimmed).

// Document WITH file_data (document_id=20951014):
{
  "document_id": "20951014",
  "url": "https://lawstronaut-prod-files-us.s3.us-east-1.amazonaws.com/...pdf?X-Amz-Expires=3600&..."
}

// Document WITHOUT file_data (document_id=27732019, HTML-only Irish Act):
{}

Notes.

• Presigned URLs expire after 3600 seconds (1 hour) — re-fetch if the user might use the link later.
• Files are served from lawstronaut-prod-files-us.s3.us-east-1.amazonaws.com.
• Always check file_data from list_documents first; calling get_source_url on a doc whose file_data is empty {} returns {} (not an error).

5.15 horizon_scan

Purpose. Mode 1, freshness-first. "What has Lawstronaut crawled / published / updated recently in this jurisdiction?"

Parameters.

Live example (trimmed).

{
  "pagination": { "total_count": 219, "limit": 3, "offset": 0 },
  "data": [{
    "document_id": 27732019,
    "title": "Protection of Employees (Employers' Insolvency) (Amendment) Act 2026",
    "issuing_authority": "Office of the Attorney General (Ireland)",
    "type_of_authority": "Act",
    "legal_link": "https://www.irishstatutebook.ie/2026/en/act/pub/0007/index.html",
    "status": "new", "version": 1,
    "crawling_date": "2026-05-20T02:39:38.557Z",
    "portal_name": "Irish Statute Book"
  }]
}

Notes.

• date_field="crawling_date" answers "what arrived in our warehouse?"; "publication_date" answers "what officially published in the period?".
• Use this instead of list_documents for time-windowed queries — it's optimized for the freshness path.
• Document object structure matches list_documents, so downstream parsers are reusable.