Google Spanner MCP Server by CData
@CDataSoftware
This project builds a read-only MCP server. For full read, write, update, delete, and action capabilities and a simplified setup, check out our free CData MCP Server for Google Spanner (beta): https://www.cdata.com/download/download.aspx?sku=OGZK-V&type=beta
Overview
What is Google Spanner MCP Server by CData?
A read-only Model Context Protocol (MCP) server that allows LLMs like Claude Desktop to query live Google Spanner data using natural language. It wraps the CData JDBC Driver for Google Spanner, which exposes Google Spanner as relational SQL models.
How to use Google Spanner MCP Server by CData?
Clone the repository, build the JAR with mvn clean install, download and license the CData JDBC Driver for Google Spanner, configure a connection string using the driver’s utility, create a .prp file with the JDBC URL and other properties, then add the server to your Claude Desktop config file using the java -jar command pointing to the JAR and .prp file. Restart Claude Desktop to see the server.
Key features of Google Spanner MCP Server by CData
- Read-only access to Google Spanner data
- Natural language querying via LLMs
- Built‑in tools:
get_tables,get_columns,run_query - Returns data in CSV format
- Uses stdio transport for local clients
- Requires CData JDBC Driver for Google Spanner
Use cases of Google Spanner MCP Server by CData
- Ask an AI assistant about Google Spanner data without writing SQL
- Explore table schemas and columns in Google Spanner
- Retrieve live business data from Google Spanner for analysis
- Integrate Google Spanner data into AI‑powered chat workflows
FAQ from Google Spanner MCP Server by CData
Is the server read-only?
Yes, this project builds a read-only MCP server that supports only SELECT queries.
What tools does the server provide?
Three tools: {servername}_get_tables (list tables), {servername}_get_columns (list columns of a table), and {servername}_run_query (execute a SQL SELECT query).
What are the runtime dependencies?
You need Java, the CData JDBC Driver for Google Spanner (installed and licensed), and the built JAR file CDataMCP-jar-with-dependencies.jar.
How is the server transported?
It uses stdio transport, so it can only be used with clients running on the same machine.
How do I authenticate to Google Spanner?
Authentication is handled by the CData JDBC Driver. The README mentions OAuth support; you will need to authenticate in your browser when testing the connection via the driver’s connection string utility.
