MCP ODBC Server

MCP ODBC Server is an MCP server that provides large language models with structured access to ODBC-compliant data sources via local Data Source Names (DSNs). Built as a TypeScript application on top of node-odbc, the server bridges Model Context Protocol clients directly to the host system's local ODBC Driver Manager. Software developers, data analysts, and database administrators use this server to interact with relational databases and semantic stores directly from LLM interfaces. The server enables agents to inspect database metadata, discover schemas, retrieve table layouts, and describe column types. It also allows clients to run arbitrary SQL queries with output formatted as JSON, Markdown tables, or JSON Lines (JSONL). In addition to standard SQL operations across generic ODBC drivers, it includes specialized support for OpenLink Virtuoso environments, enabling hybrid SPASQL queries, native SPARQL queries against RDF graphs, and automated interactions with Virtuoso AI assistant procedures.

Category: Data & Analytics

Tags: database-connectivity, dsn, odbc, SQL

Visit MCP ODBC Server

How to install and configure MCP ODBC Server

1. Ensure Node.js version 21.1.0 or higher is installed and verify your unixODBC environment using odbcinst -j and odbcinst -q -s. 2. Clone the repository and change directory: sh git clone https://github.com/OpenLinkSoftware/mcp-odbc-server.git cd mcp-odbc-server 3. Install the required dependencies: sh npm install @modelcontextprotocol/sdk zod tsx odbc dotenv 4. Configure your .env file with your connection parameters such as ODBC_DSN, ODBC_USER, ODBC_PASSWORD, and ODBCINI. 5. Register the server in your MCP client configuration using the tsx executable pointing to ./src/main.ts: json { "mcpServers": { "odbc": { "command": "npx", "args": ["tsx", "/path/to/mcp-odbc-server/src/main.ts"], "env": { "ODBCINI": "/Library/ODBC/odbc.ini" } } } }

What you can do with MCP ODBC Server

• Discover and explore database architectures by listing accessible schemas, filtering table names, and inspecting detailed column schemas. - Execute SQL queries from an MCP client to retrieve structured records formatted as JSON Lines or Markdown tables. - Run native SPARQL queries through OpenLink Virtuoso to inspect RDF knowledge graphs and extract semantic triples directly. - Execute hybrid SPASQL queries combining relational SQL data with graph-based SPARQL lookups in unified database calls. - Interact with the Virtuoso AI Support Assistant by forwarding prompts and credentials through dedicated MCP database tools.

Key facts

• https://github.com/OpenLinkSoftware/mcp-odbc-server
• Data & Analytics, Databases & Data Stores
• database-connectivity, dsn, odbc, SQL

Part of MCP Servers

What is MCP ODBC Server?

MCP ODBC Server is a TypeScript Model Context Protocol server that uses node-odbc to route database commands from LLM clients to local ODBC Driver Managers. It allows AI tools to inspect schemas, read table structures, and run queries against any database with an active DSN.

How do I install MCP ODBC Server?

Clone the repository from GitHub, verify that Node.js 21.1.0 or higher is active, and install packages using npm install. Set up your ODBC configuration paths and credentials in an environment file, then point your MCP client to the server entry point using npx and tsx.

What tools are included with MCP ODBC Server?

The server includes schema discovery tools like get_schemas, get_tables, filter_table_names, and describe_table. Query execution tools include query_database, execute_query, and execute_query_md. For Virtuoso databases, it provides spasql_query, sparql_query, and virtuoso_support_ai.

Which databases work with MCP ODBC Server?

Any database management system accessible through an installed ODBC driver and configured DSN can work with this server. While tested and documented using OpenLink Virtuoso, generic SQL querying and schema inspection functions operate across any standard ODBC connector.