apidriver_get_tablesLABSCO SUMMARY
The licensed driver does the hard part.
Three tools is the whole surface, and that is the point — the JDBC driver turns API Driver into ordinary tables, so the model writes SELECT statements instead of learning an API shape. What you trade for that is a driver install and a licensing step before the first query, and a server that only reads: there is no write, update, or delete path here.
What it is
A local Java server that puts API Driver behind three SQL tools. CData's JDBC Driver for API Driver exposes the service as relational tables, and this server wraps the driver so a model can find a table, read its columns, and run a SELECT against live data without anyone writing the API calls.
What you get
apidriver_get_tablesreturns the tables the driver exposes, as CSV with a header rowapidriver_get_columnsreturns the columns on one table, in the same CSV shapeapidriver_run_queryexecutes a SQL SELECT statement- The
apidriverpart is thePrefixyou set in the property file (Prefix=apidriverin the sample), so several CData servers can sit in one client without their tool names colliding - A
Tablessetting that only picks which tables also get a ready-made description (blank means none); it does not limit what the tools can list or query, so limit the connection or the account it signs in with
Requirements
Build it yourself: clone the repository and run mvn clean install, which produces CDataMCP-jar-with-dependencies.jar. Install the CData JDBC Driver for API Driver separately and license it — java -jar cdata.jdbc.apidriver.jar --license, then name, email, and TRIAL or your key. Write a api-driver.prp file carrying DriverClass=cdata.jdbc.apidriver.APIDriverDriver and the JDBC URL from the driver's connection-string utility, and launch the server as java -jar CDataMCP-jar-with-dependencies.jar api-driver.prp. Transport is stdio, so the server runs on the same machine as the client.
Setup effort
Build from source — clone the repository, then point your client at the binary
USE CASES
01
Answering a question about live API Driver data
Ask in plain language; the model locates the table, checks its columns, then writes the SELECT itself.
02
Finding out what is actually available
List the tables and their columns first, so you know what can be asked before you ask it.
03
Running several CData sources in one client
Give each property file its own Prefix so the tool names stay distinct across servers.
3 tools
Listed in project docs. Your client may show a different list.
apidriver_get_columnsapidriver_run_queryBEFORE YOU INSTALL
What to know first
Tools you get
3 tools named in the project's documentation. Tool names take the Prefix= value from the .prp file, such as apidriver_get_tables. All tools are read-only; writes need CData's paid MCP Server. The Tables setting does not limit what the tools reach; your connection's access does.
Where it runs
Both documented setups run the server on your machine.
INSTALLATION
Connect your client
Clone the repository, then point your client at the binary. This one is built from source rather than installed with a single command.
Server setup
TERMINAL
git clone https://github.com/cdatasoftware/api-driver-mcp-server-by-cdata.git
cd api-driver-mcp-server-by-cdata
mvn clean installDownload and install the CData JDBC Driver for API Driver from https://www.cdata.com/drivers/apidriver/download/jdbc
License the driver: in the driver's lib folder run java -jar cdata.jdbc.apidriver.jar --license, then enter name, email and "TRIAL" (or your license key)
Run java -jar cdata.jdbc.apidriver.jar to open the Connection String utility, test the connection and copy the connection string
Create api-driver.prp with:
FILE · api-driver.prp
Prefix=apidriver
ServerName=CDataAPIDriver
ServerVersion=1.0
DriverPath=<YOUR_DRIVER_PATH>
DriverClass=cdata.jdbc.apidriver.APIDriverDriver
JdbcUrl=<YOUR_JDBC_URL>
Tables=Tables only picks which tables also get a ready-made description for your assistant, and empty means none; the server can still list and query every table your connection string reaches, so to keep data out of reach, limit that connection or the account it signs in with.
Add the mcpServers entry below to claude_desktop_config.json and restart the client
MCP Clients
Claude Desktop — Linux, macOS, Windows
JSON
{
"mcpServers": {
"api-driver": {
"command": "<YOUR_PATH_TO_JAVA>",
"args": [
"-jar",
"<YOUR_PATH_TO_CDATA_MCP_JAR_WITH>",
"<YOUR_PATH_TO_API_DRIVER>"
]
}
}
}