Yhteentoimivuusalusta MCP Server
Enables searching, validating, and referencing Finnish standardized terminology, data models, and code lists to ensure consistency in design documentation for Finnish government and building industry projects.
README
Yhteentoimivuusalusta MCP Server
An MCP (Model Context Protocol) server that integrates with Finland's Yhteentoimivuusalusta (Interoperability Platform) to assist with design documentation by ensuring proper use of standardized Finnish terminology, data models, and code values.
Overview
This tool helps when writing design documents, technical specifications, or architecture descriptions for Finnish government or building industry projects by:
- Suggesting correct standardized terms from relevant vocabularies
- Referencing appropriate data models for data structure descriptions
- Providing valid code list values for enumerations
- Validating terminology consistency across documentation
Warning
This program is coded, poorly, by someone who has absolutely no idea what they are doing. The code is also heavily AI assisted. Use at your own risk and if possible, via Docker.
Supported Platforms
| Platform | URL | Content |
|---|---|---|
| Sanastot | sanastot.suomi.fi | ~90 terminologies and vocabularies |
| Tietomallit | tietomallit.suomi.fi | ~170 data models and schemas |
| Koodistot | koodistot.suomi.fi | 700+ code lists and classifications |
Installation
Prerequisites
- Python 3.11 or higher
- pip or uv package manager
- Docker (optional, recommended for sandboxed execution)
Option 1: Docker (Recommended)
Docker provides sandboxed execution with minimal attack surface:
# Clone and build
git clone https://github.com/yaskael/yhteentoimivuusagentti.git
cd yhteentoimivuusagentti
docker build -t yhteentoimivuusalusta-mcp:latest .
# Test the build
docker run --rm yhteentoimivuusalusta-mcp:latest python -c "import yhteentoimivuusalusta_mcp.server; print('OK')"
See docs/DOCKER.md for detailed Docker setup instructions.
Option 2: Install from source
# Clone the repository
git clone https://github.com/yaskael/yhteentoimivuusagentti.git
cd yhteentoimivuusagentti
# Install in development mode
pip install -e .
# Or with uv
uv pip install -e .
Optional dependencies
# For improved fuzzy matching
pip install rapidfuzz
# For Finnish NLP (lemmatization)
pip install voikko
Configuration
Copy the example configuration file and adjust as needed:
cp config.yaml.example config.yaml
Configuration options
# config.yaml
cache:
enabled: true
directory: ~/.cache/yhteentoimivuusalusta
rate_limit:
requests_per_second: 10.0
# API endpoints (defaults shown)
apis:
sanastot:
base_url: https://sanastot.suomi.fi/terminology-api
tietomallit:
base_url: https://tietomallit.suomi.fi/datamodel-api
koodistot:
base_url: https://koodistot.suomi.fi/codelist-api/api/v1
Usage
Running the server
python -m yhteentoimivuusalusta_mcp.server
Claude Desktop Integration
Option 1: Automatic Setup (Recommended)
Run the setup script to automatically configure Claude Desktop:
python scripts/setup_claude_desktop.py
This will detect your OS, find the correct config location, and add the MCP server configuration.
To preview without writing:
python scripts/setup_claude_desktop.py --print-only
Option 2: Manual Configuration
Add to your Claude Desktop configuration (claude_desktop_config.json):
| OS | Config Location |
|---|---|
| macOS | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Windows | %APPDATA%\Claude\claude_desktop_config.json |
| Linux | ~/.config/Claude/claude_desktop_config.json |
{
"mcpServers": {
"yhteentoimivuusalusta": {
"command": "python",
"args": ["-m", "yhteentoimivuusalusta_mcp.server"],
"cwd": "/path/to/yhteentoimivuusagentti"
}
}
}
Option 3: Docker (Sandboxed)
For enhanced security, run the MCP server in a Docker container:
{
"mcpServers": {
"yhteentoimivuusalusta": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"--memory=512m", "--cpus=1.0",
"-v", "yhteentoimivuusalusta-cache:/app/cache",
"yhteentoimivuusalusta-mcp:latest"
]
}
}
}
This provides:
- Isolated filesystem (no access to host files)
- Memory and CPU limits
- Non-root execution
- Persistent cache via Docker volume
Option 4: Install as Desktop Extension (MCPB)
This project includes a manifest.json (v0.3) for MCPB desktop extensions:
# Create the bundled extension
python scripts/create_bundle.py
# This creates:
# - build/yhteentoimivuusalusta-0.1.0.mcpb (bundled)
# - build/unpacked/ (for unpacked installation)
To install the extension:
-
Bundled (.mcpb): Double-click or drag
build/yhteentoimivuusalusta-0.1.0.mcpbto Claude Desktop -
Unpacked (development): In Claude Desktop settings, add the extension pointing to
build/unpacked/directory -
Using MCPB CLI:
npm install -g @anthropic-ai/mcpb mcpb install ./build/yhteentoimivuusalusta-0.1.0.mcpb
After installation, restart Claude Desktop.
Available Tools
The server provides 13 MCP tools across three categories:
Terminology Tools (Sanastot)
| Tool | Description |
|---|---|
search_terminology |
Search for terms across vocabularies with fuzzy matching |
get_concept_details |
Get detailed information about a specific concept |
list_vocabularies |
List all available vocabularies |
Data Model Tools (Tietomallit)
| Tool | Description |
|---|---|
search_datamodel |
Search for data models by name or description |
get_datamodel_classes |
Get all classes from a specific data model |
get_model_vocabulary_links |
Find vocabularies linked to a data model |
Code List Tools (Koodistot)
| Tool | Description |
|---|---|
search_codelist |
Search for code lists by name or description |
get_codes |
Get all codes from a specific code list |
export_codes_csv |
Export codes to CSV format |
Cross-Platform Tools
| Tool | Description |
|---|---|
validate_terminology |
Validate text against standardized vocabularies |
unified_search |
Search all three platforms simultaneously |
suggest_references |
Analyze text and suggest relevant standards to reference |
get_codelist_for_attribute |
Find appropriate code lists for data model attributes |
Examples
Search for a term
Tool: search_terminology
Arguments: { "query": "rakennus", "vocabulary_id": "rakymp" }
Validate terminology in a document
Tool: validate_terminology
Arguments: {
"text": "Rakennuksen kerrosala lasketaan ulkoseinien ulkopintojen mukaan.",
"vocabulary_ids": ["rakymp"]
}
Unified search across all platforms
Tool: unified_search
Arguments: {
"query": "kaava",
"platforms": ["sanastot", "tietomallit", "koodistot"],
"limit": 5
}
Suggest standards for documentation
Tool: suggest_references
Arguments: {
"text": "Järjestelmä käsittelee rakennuslupahakemuksia ja niiden liitteitä.",
"include_vocabularies": true,
"include_datamodels": true,
"include_codelists": true
}
Common Resource IDs
Vocabularies (Sanastot)
rakymp- Built environment vocabulary (Rakennetun ympäristön sanasto)jhka- Public administration architecture (Julkisen hallinnon kokonaisarkkitehtuuri)oksa- Education vocabulary (Opetus- ja koulutussanasto)kela- Social security vocabulary (Kelan sanasto)
Data Models (Tietomallit)
rytj-kaava- Spatial planning data modelraktkk- Physical building data modeldigione- Education data models
Code Registries (Koodistot)
rakennustieto- Building information codeskoulutus- Education codesjulkishallinto- Public administration codes
Features
Caching
- Persistent disk cache with configurable TTL
- Automatic cache invalidation
- Offline mode support with stale cache fallback
Performance
- Rate limiting (token bucket algorithm, 10 req/sec default)
- Parallel API requests where possible
- Response caching to minimize API calls
Reliability
- Automatic retry with exponential backoff
- Graceful degradation when APIs are unavailable
- Offline mode returns cached data when network fails
Project Structure
yhteentoimivuusalusta-mcp/
├── src/
│ └── yhteentoimivuusalusta_mcp/
│ ├── __init__.py
│ ├── server.py # MCP server entry point
│ ├── tools/
│ │ ├── terminology.py # Sanastot tools
│ │ ├── datamodel.py # Tietomallit tools
│ │ ├── codelist.py # Koodistot tools
│ │ ├── validation.py # Text validation
│ │ └── unified.py # Cross-platform tools
│ ├── clients/
│ │ ├── base.py # Base HTTP client
│ │ ├── sanastot.py # Sanastot API client
│ │ ├── tietomallit.py # Tietomallit API client
│ │ └── koodistot.py # Koodistot API client
│ ├── models/
│ │ └── schemas.py # Pydantic models
│ └── utils/
│ ├── cache.py # Caching utilities
│ ├── config.py # Configuration loader
│ └── fuzzy.py # Fuzzy matching
├── docs/
│ └── DOCKER.md # Docker setup guide
├── tests/
├── Dockerfile # Docker build configuration
├── docker-compose.yml # Docker Compose configuration
├── pyproject.toml
└── config.yaml.example
Development
Running tests
pytest tests/
Code formatting
ruff format src/ tests/
ruff check src/ tests/
License
MIT License
References
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.