MCP Server Jdbc

What quarkiverse/mcp-server-jdbc MCP server does

The quarkiverse/mcp-server-jdbc MCP server gives MCP-enabled AI applications access to a database through JDBC. It is intended for use with JDBC-compatible systems rather than one specific database vendor, so the same server can target databases such as PostgreSQL, MySQL, Oracle, and SQLite.

Database operations include querying data and changing stored records. The available use cases described for the server include inserting, updating, and deleting data, as well as storing and retrieving information. This places it in the 🗄️ Databases category for agent workflows that need direct database interaction.

How it works

The server uses a JDBC URL to identify the database connection target. An MCP client connects to the running server, and the connected AI application can then request supported database operations through MCP. The repository implements its servers in Java with the Quarkus MCP server framework.

The project presents the JDBC server as a standalone entry point named jdbc. It can be launched with JBang, which makes the server available without requiring a separate application build from the repository for the documented quick-start path.

Setup and configuration

Start the server with:

jbang jdbc@quarkiverse/quarkus-mcp-servers

The documented server model uses a JDBC URL to connect to the selected database. The supplied material does not specify additional environment variables, connection properties, authentication settings, or database-driver installation steps, so those details should be confirmed in the JDBC server's own README before deployment.

Where the command uses jbang, the project documentation says it can be replaced with one of these alternatives:

• npx @jbangdev/jbang
• uvx jbang
• pipx jbang

The server can be used with MCP-enabled applications, including Claude Desktop, which the repository explicitly names as an example client.

Tools and capabilities

The quarkiverse/mcp-server-jdbc MCP server supports database actions described at a capability level rather than by individually named MCP tool identifiers in the supplied material:

• Query database data
• Insert records or other data
• Update stored data
• Delete stored data
• Store and retrieve information through a JDBC connection

Because the connection is based on JDBC, the relevant database is selected through its JDBC URL. The repository lists PostgreSQL, MySQL, Oracle, and SQLite as examples of supported database types.

Limitations and notes

The available documentation does not list the exact MCP tool names, request schemas, result formats, JDBC driver versions, or security controls. It also does not document transaction behavior, schema discovery, connection pooling, or read-only safeguards. Treat those areas as implementation details to verify before allowing an agent to modify production data.

The server's ability to connect to a particular database still depends on that database being JDBC-compatible and on the runtime having the connection details and supporting requirements it needs. The repository is part of a broader collection of Java and Quarkus MCP servers, so instructions for other entries in that collection should not be assumed to apply to this JDBC server.