XBase

You run a read-only MCP server that makes live xBase data accessible to language models through a simple MCP interface. By wrapping the CData JDBC Driver for xBase, you can ask natural-language questions and retrieve up-to-date data without writing SQL. This guide shows you how to install, configure, and use the MCP server end-to-end.

How to use

Once you have the MCP server running, you can interact with it using an MCP client (such as Claude Desktop) to read data from xBase sources. You do not need to write SQL; instead, you ask questions in natural language and the client uses built-in tools to discover tables, examine columns, and run read-only queries. Typical use cases include asking for summaries of data, checking counts or statuses, and retrieving specific records or aggregates from the xBase data exposed by the JDBC driver.

How to install

Prerequisites: you need a Java Runtime Environment and Maven installed to build and run the MCP server.

# 1) Clone the repository
git clone https://github.com/cdatasoftware/xbase-mcp-server-by-cdata.git
cd xbase-mcp-server-by-cdata

# 2) Build the server
mvn clean install

# This creates the JAR file: CDataMCP-jar-with-dependencies.jar

3. Download and install the CData JDBC Driver for xBase from the following URL and follow the installer prompts: https://www.cdata.com/drivers/xbase/download/jdbc

4. License the CData JDBC Driver. In your installation directory, locate the lib folder and run: java -jar cdata.jdbc.xbase.jar --license Enter your name, email, and the word TRIAL (or your license key) when prompted.

5. Configure your connection to the data source (the example uses Salesforce). Run the connection string utility to build and test the connection string, then copy it for later use: java -jar cdata.jdbc.xbase.jar Use the utility to configure the connection, test the connection, and copy the JDBC URL string. If OAuth is used, authenticate in your browser during the test.

6. Create a .prp configuration file for your JDBC connection. Save it as xbase.prp with the following properties. You can copy the values from your connection string test if needed:

ServerName=CDataxBase
ServerVersion=1.0
DriverPath=PATH\TO\cdata.jdbc.xbase.jar
DriverClass=cdata.jdbc.xbase.xBaseDriver
JdbcUrl=jdbc:xbase:InitiateOAuth=GETANDREFRESH;
Tables=```
This file wires the MCP server to the xBase data source.

## Additional setup for Claude Desktop configuration

To let Claude Desktop know about the new MCP server, create or update claude\_desktop\_config.json with an entry that points to the MCP JAR and the xbase.prp file. Use the following structure for Windows, Linux, or macOS as shown.

{ "mcpServers": { "xbase_mcp": { "command": "PATH\TO\java.exe", "args": [ "-jar", "PATH\TO\CDataMCP-jar-with-dependencies.jar", "PATH\TO\xbase.prp" ] }, ... } }

## Running the server

Start the MCP server on its own using the Java JAR and the xbase.prp configuration. The following command runs the server with the specified PRP file. Note that the server uses stdio and runs on the same machine as the client.

java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/Salesforce.prp

Note: The example uses Salesforce.prp in the run command to illustrate the startup flow. You should point to your actual PRP configuration file (xbase.prp) when you start the server.

Usage details

After configuration, you can let the AI client access the built-in tools to read data. You don’t need to call the tools directly; simply ask questions about the underlying data system, such as summarizing data, counting records, or retrieving specific fields.

Tools & Descriptions

In the definitions below, {servername} refers to the MCP server name you configured. The tools available are:

{servername}_get_tables - Retrieves a list of tables available in the data source. The output is in CSV format with column headers on the first line.

{servername}_get_columns - Retrieves a list of columns for a given table. The output is in CSV format with column headers on the first line.

{servername}_run_query - Executes a SQL SELECT query against the data source.

Troubleshooting

If you cannot see your MCP server in Claude Desktop, fully quit the client and restart. If Claude Desktop cannot retrieve data, verify the connection string in the .prp file using the connection string builder and ensure it is copied into the PRP correctly. If you have trouble connecting, contact CData Support. For further feedback, join the CData Community.

License

This MCP server is licensed under the MIT License. You may use, modify, and distribute it in accordance with the MIT License terms.

All Supported Sources

The MCP server supports a wide range of data sources, including xBase, Salesforce, Google Drive, S3, Snowflake, SharePoint, Dynamics, SAP, Oracle, MySQL, PostgreSQL, and many more. The list below shows a portion of the sources available in this ecosystem.

Available tools

xbase_get_tables

Retrieves a list of tables available in the xBase data source. Output is CSV with column headers on the first line.

xbase_get_columns

Retrieves a list of columns for a specified table. Output is CSV with column headers on the first line.

xbase_run_query

Executes a SQL SELECT query against the xBase data source.