hu-eli-mcp
Enables AI agents to query Hungary's official legislation database (NJT) using native ELI identifiers, retrieving metadata and full text with verifiable citations.
README
hu-eli-mcp
<!-- mcp-name: io.github.matematicsolutions/hu-eli-mcp -->
An MCP server for Hungary's official national legislation database, the Nemzeti
Jogszabalytar (NJT, njt.jog.gov.hu), maintained by MKIFK. It gives an AI agent a piece of
Hungarian legislation resolved by its native ELI (European Legislation Identifier), with a
verifiable citation: the ELI itself, a human-readable citation, and a link to the official
source.
Part of the eu-legal-mcp line by MateMatic — one connector per EU member state, the same citation contract everywhere.
On ELI. Unlike the Netherlands and Sweden (earlier connectors in this line, neither of which publishes native ELI), Hungary has a genuine, documented national ELI implementation since 2023 (
njt.jog.gov.hu/eli/urisemak).eli_uriis therefore the real/eli/URI the tool resolved — never a substitute identifier. SeeDISCOVERY.md.
Tools
| Tool | What it does |
|---|---|
hu_get_act(year, serial) |
Metadata for a Hungarian Act (torveny) by ELI, e.g. year 2013, serial 5 (the Civil Code). No issuer code needed. |
hu_get_legislation(doc_type, year, issuer, serial) |
Metadata for any other legislation (decrees etc.) by ELI — type, year, issuer (case-sensitive) and serial. |
hu_get_text(doc_type, year, serial, issuer=None) |
The full document HTML content for either an Act or other legislation. |
hu_list_doc_types() |
Valid NJT document-type codes (e.g. TV=Act, R=decree). |
hu_list_issuers() |
Valid NJT issuer codes (e.g. Korm=Government) — note the exact casing matters. |
Every response carries the citation contract:
eli_uri— the native ELI URI resolved, e.g.https://njt.jog.gov.hu/eli/TV/2013/5.human_readable_citation— built from the parsed title + subtitle, e.g. "2013. évi V. törvény (a Polgári Törvénykönyvről)".source_url— the browsablenjt.jog.gov.hu/jogszabaly/...document page.
No free-text search
NJT has no REST/JSON API and its search/browse UI is a JavaScript SPA requiring a session
(not usable headlessly). Documents must be addressed by a known ELI — use
hu_list_doc_types / hu_list_issuers to discover valid codes, then hu_get_act /
hu_get_legislation. An unresolvable ELI (wrong year/serial/issuer, or wrong issuer-code
case) surfaces as not_found — NJT itself redirects it to a search-assist page rather than
a 404.
Install
pip install -e ".[dev]"
Register it with your MCP client (see .mcp.json.example):
{
"mcpServers": {
"hu-eli-mcp": {
"command": "hu-eli-mcp",
"env": {
"HU_ELI_BASE_URL": "https://njt.jog.gov.hu",
"HU_ELI_CACHE_DIR": "~/.matematic/cache/hu-eli",
"HU_ELI_AUDIT_DIR": "~/.matematic/audit"
}
}
}
}
Design
- Public data only. Read-only against the keyless, official
njt.jog.gov.hu; nothing is sent beyond the ELI path being resolved. Honorsrobots.txt(never calls/search/*). - Audit log. Every call appends one JSON line to
~/.matematic/audit/hu-eli-mcp.jsonl(AI Act art. 12 record-keeping). - Vendor-neutral. No LLM provider, no telemetry; own backoff + on-disk cache.
- No fabrication. The ELI, title and citation are parsed from the resolved document page. If NJT's markup changes, the connector fails loudly rather than returning stale or invented data.
See CONSTITUTION.md (the 4 principles) and DISCOVERY.md (how the source was mapped,
including the njt.hu → njt.jog.gov.hu domain discovery and the issuer-code casing gotcha).
Tests
pytest tests/test_instructions_drift.py # offline
pytest tests/test_smoke.py # live NJT
Licence
Apache-2.0. The Hungarian legislation served is official public data of Hungary; this connector adds no rights over it.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.