tally -by-cdata

This MCP server provides a read-only interface to Tally data via the MCP (Model Context Protocol). It wraps the CData JDBC Driver for Tally and exposes a small set of tools that let an AI client discover available data (tables and columns) and run read-only queries without writing SQL directly. Typical usage is to configure Claude Desktop or a compatible MCP client to reach the server, then issue the built-in tools named after the server (e.g., tally_get_tables, tally_get_columns, tally_run_query). The tools return results in CSV for tabular data and can be invoked via JSON-RPC payloads as shown in the examples. You do not need to manually craft SQL in normal interactions; the client can request data explorations and simple selects through the exposed tools.

Prerequisites:

• Java Runtime / JDK installed (the server runs as a Java JAR).
• Maven (for building the server JAR from source).
• CData JDBC Driver for Tally installed and licensed, if you plan to test with the JDBC driver locally.

Installation steps:

1. Build the MCP server JAR from source:

git clone https://github.com/cdatasoftware/tally-mcp-server-by-cdata.git
cd tally-mcp-server-by-cdata
mvn clean install

This produces a JAR named similar to CDataMCP-jar-with-dependencies.jar (location usually in target/).

2. Obtain and license the CData JDBC Driver for Tally:

• Download from https://www.cdata.com/drivers/tally/download/jdbc
• Follow the licensing steps provided by CData to license the driver (typically via a jar command and the license prompt).

3. Prepare a .prp connection file for your JDBC setup (example named tally.prp) with the required properties (Prefix, ServerName, ServerVersion, DriverPath, DriverClass, JdbcUrl, Tables). Ensure the DriverPath points to the CData JDBC JAR and JdbcUrl contains a valid connection string.

4. Run the MCP server:

java -jar /path/to/CDataMCP-jar-with-dependencies.jar /path/to/tally.prp

Note: The MCP server uses stdio and is intended to be used by clients on the same machine.

Tips and notes:

• This is a read-only MCP server. If you need full read/write/update/delete capabilities, use the full CData MCP Server for Tally.
• The server operates over stdio, so ensure your client runs on the same host.
• Use tally_get_tables to discover available data, tally_get_columns to list columns for a specific table, and tally_run_query to execute a read-only SELECT against the data source.
• If you encounter connection or licensing issues, verify the JdbcUrl/DriverPath in tally.prp and ensure the CData JDBC Driver is properly licensed.
• When updating the client configuration (Claude Desktop or equivalent), restart the client to pick up new MCP server definitions.