orionbelt-analytics

Overview

OrionBelt Analytics is a comprehensive MCP server that transforms relational database schemas into RDF/OWL ontologies with embedded SQL mappings. It supports multiple database systems including PostgreSQL, Snowflake, ClickHouse, Dremio, BigQuery, DuckDB, Databricks, and MySQL. The server provides relationship-aware Text-to-SQL with automatic fan-trap prevention, GraphRAG for intelligent schema discovery, and interactive charting capabilities. Its key feature is OBQC (Ontology-Based Query Check), a deterministic SQL validator that catches structural errors before queries reach the database.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• connect_database

  Connect to any supported database using .env credentials

  Try:Connect to PostgreSQL databaseTry:Connect to Snowflake warehouse
• list_schemas

  List available schemas in the connected database

  Try:Show all schemas in the databaseTry:List available schemas
• discover_schema

  Analyze schema structure with automatic GraphRAG + ontology generation

  Try:Discover schema for public schemaTry:Analyze database structure
• get_table_details

  Get detailed column, key, and constraint info for a specific table

  Try:Show details for customers tableTry:Get schema for orders table
• generate_ontology

  Generate RDF/OWL ontology from schema with SQL mapping annotations

  Try:Generate ontology for current schemaTry:Create OWL model from database
• suggest_semantic_names

  Detect abbreviations and cryptic names for business-friendly renaming

  Try:Suggest better names for columnsTry:Find cryptic identifiers to rename
• sample_table_data

  Preview table data with row limit and injection protection

  Try:Show first 10 rows of productsTry:Sample data from sales table
• execute_sql_query

  Execute SQL with OBQC validation, security checks, and fan-trap detection

  Try:Run SELECT query with validationTry:Execute sales analysis query
• generate_chart

  Generate Plotly charts (bar, line, scatter, heatmap) with MCP-UI rendering

  Try:Create bar chart of monthly salesTry:Generate line chart of user growth
• graphrag_search

  Semantic search + schema overview (auto-initialized by discover_schema)

  Try:Search for customer-related tablesTry:Find all tables with order data
• graphrag_find_join_path

  Discover join paths between tables via graph traversal

  Try:Find path between customers and ordersTry:Show join path between products and sales
• query_sparql

  Execute SPARQL queries (SELECT, ASK, CONSTRUCT — auto-detected)

  Try:Query ontology for all customersTry:Find all product categories in RDF store

Comparable tools

Installation

git clone https://github.com/ralfbecher/orionbelt-analytics
cd orionbelt-analytics
uv sync

cp .env.template .env

uv run server.py

{
  "mcpServers": {
    "OrionBelt-Analytics": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:9000/mcp",
        "--transport",
        "http-only"
      ]
    }
  }
}

FAQ

What databases does OrionBelt Analytics support?

OrionBelt Analytics supports PostgreSQL, MySQL, Snowflake, ClickHouse, Dremio, BigQuery, DuckDB/MotherDuck, and Databricks SQL.

How does OBQC work?

OBQC (Ontology-Based Query Check) is a deterministic validator that parses SQL against the loaded RDF/OWL ontology to check table/column existence, join validity, type compatibility, aggregation correctness, and fan-traps before query execution.

Compare orionbelt-analytics with