Estonia AI Kit v1.2.0 - Major MCP Server Improvements

🏰 Estonia AI Kit v1.2.0 - Major MCP Server Improvements

🚀 What's New

This release brings significant improvements across all MCP servers, introduces a new RAG system for Estonian legal documents, and provides more accurate package status information.

📦 Package Updates

🔶 RIK MCP Server v1.2.0 (WIP)

• SQLite Database Support: Local database for offline company and board member queries
• Data Import Scripts: Automated tools to build and update database from official sources
• Enhanced Performance: Database queries execute in <50ms vs 2-3s for file parsing
• Comprehensive Dataset: Partial dataset of Estonian companies and board members
• Streaming Import: Efficient memory usage for large datasets

🔶 Open Data MCP Server v1.2.0 (WIP)

• Curated Tables: Pre-defined known tables for common searches
• Enhanced Search: Keyword-based filtering with categories (agriculture, population, economy, environment)
• Better Error Handling: Alternative suggestions when API calls fail
• Statistics Estonia Integration: Direct API integration with fallback mechanisms

🔴 EMTA MCP Server v1.2.0 (POC)

• Proof of Concept: Currently uses mock data for demonstration
• Server Improvements: Enhanced startup/shutdown handling and error management
• X-Road Ready: Architecture prepared for X-Road integration (authentication required)
• Process Management: Better signal handling and graceful shutdown

✅ Shared Utilities v1.2.0 (Ready)

• Type Definitions: Enhanced TypeScript types across all packages
• Common Utilities: Shared functions for X-Road and API interactions

🆕 New RAG System

🔶 Riigiteataja Legal Document RAG (WIP)

• Vector Search: Estonian legal document analysis and search
• Document Scraping: Automated extraction from Riigiteataja
• FAISS Integration: High-performance vector similarity search
• TypeScript Implementation: Modern tooling with comprehensive type safety

🛠️ Technical Improvements

• Better Server Lifecycle: Improved startup/shutdown across all MCP servers
• Enhanced Error Handling: More robust error reporting and recovery
• Modular Architecture: Exportable main functions for better testing
• Performance Optimizations: Reduced memory usage and faster query responses

📋 Installation & Configuration

Prerequisites

# Clone repository and install dependencies
git clone https://github.com/stefanoamorelli/estonia-ai-kit.git
cd estonia-ai-kit
npm install
npm run build

Claude Desktop Configuration

{
  "mcpServers": {
    "estonia-rik": {
      "command": "node",
      "args": ["/path/to/estonia-ai-kit/mcp/rik/dist/index.js"]
    },
    "estonia-emta": {
      "command": "node",
      "args": ["/path/to/estonia-ai-kit/mcp/emta/dist/index.js"]
    },
    "estonia-open-data": {
      "command": "node",
      "args": ["/path/to/estonia-ai-kit/mcp/open-data/dist/index.js"]
    }
  }
}

🎯 Package Status Legend

• 🔴 POC (Proof of Concept): Mock data, not production-ready
• 🔶 WIP (Work in Progress): Partial functionality, actively developed
• ✅ Ready: Full functionality, production-ready

📝 Documentation

• Main README: README.md
• RIK Package: mcp/rik/README.md
• EMTA Package: mcp/emta/README.md
• Open Data Package: mcp/open-data/README.md
• RAG System: rag/riigiteataja/README.md

🤝 Contributing

This is a community-driven project. Contributions, bug reports, and feature requests are welcome!

⚖️ License

Licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).

---

Note: This is experimental software for development and educational purposes. Always verify data accuracy for production use cases.