airtable -by-cdata

This MCP server provides a read-only interface to Airtable data via the CData MCP wrapper around the CData JDBC Driver for Airtable. Once configured, you can query live Airtable data through the MCP tools exposed by the server without writing SQL directly in your client. The server exposes a small set of tools designed for downstream AI clients to discover schema and retrieve data: get_tables to enumerate available tables, get_columns to inspect columns within a table, and run_query to execute a SQL SELECT against the underlying Airtable data source. In practice, you connect to the MCP server from your MCP client (such as Claude Desktop) and invoke these tools by name, receiving results in CSV or tabular formats suitable for natural language processing pipelines. The setup emphasizes local, read-only access to Airtable via the JDBC driver, enabling live data queries while preventing modification of source data.

To use the tools, first ensure the MCP server is running with the proper Airtable.prp configuration. Then, from your client, call airtable_by_cdata_get_tables to get the list of accessible tables, followed by airtable_by_cdata_get_columns for specific tables, and finally airtable_by_cdata_run_query to perform read-only queries like SELECT * FROM [Table] WHERE ... . The output formats are designed for easy consumption by LLMs: tables are returned in CSV with column headers on the first line.

Prerequisites:

• Java Runtime (recommended JRE/JDK 8+)
• Maven (for building the MCP server)
• Airtable JDBC Driver installed and licensed according to the CData instructions

Step-by-step installation:

1. Install prerequisites:

   • Install Java: download from Oracle/OpenJDK and ensure java -version works in your shell
   • Install Maven: follow your OS instructions to install Maven and verify mvn -v

2. Build the MCP server:

   • git clone https://github.com/cdatasoftware/airtable-mcp-server-by-cdata.git
   • cd airtable-mcp-server-by-cdata
   • mvn clean install
   • This will produce a JAR named similar to CDataMCP-jar-with-dependencies.jar in the target directory.

3. Install and license the CData JDBC Driver for Airtable (as described in the README):

   • Download the JDBC driver from the CData site and install it according to platform instructions
   • In the driver’s lib directory, run the license command and provide your name, email, and license key (or TRIAL)

4. Prepare the prp configuration file for Airtable (airtable.prp) using the properties shown in the README, including DriverPath, DriverClass, and JdbcUrl, tailored to your Airtable setup.

5. Run the MCP server:

   • java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/airtable.prp
   • The server will start and listen for MCP client requests on STDIO (local machine only) as described in the README.

Tips and known considerations:

• This server is read-only by default; if you need write/update/delete capabilities, use the full MCP server variant described in the README under the Airtable MCP Server options.
• The server uses stdio, so it must be interacted with from clients running on the same machine.
• Ensure your Airtable connection string (JdbcUrl) is tested and the CData driver is properly licensed before starting the MCP server.
• If Claude Desktop or other clients do not detect the MCP server, fully quit and restart the client so it rescans the MCP configuration. Ensure the claude_desktop_config.json (or equivalent) includes the correct path to the jar and prp file.
• The tool names are derived from the server name in the config; in examples they follow the pattern {servername}_get_tables, {servername}_get_columns, and {servername}_run_query. Replace {servername} with your configured server name (e.g., airtable_by_cdata).