veeva -by-cdata

This MCP server exposes a read-only interface to Veeva data via the MCP protocol. It wraps the CData JDBC Driver for Veeva and presents the data as relational models accessible through the MCP tools. Clients (such as Claude Desktop or other MCP-enabled LLM clients) can query metadata, list available tables and columns, and execute read-only SQL-like queries through the provided tools. The available tools follow a naming convention based on the server name (for example veeva_get_tables, veeva_get_columns, and veeva_run_query). You typically do not need to call these tools directly; you can simply ask the AI to retrieve information and the client will route requests to the appropriate tools behind the scenes.

Prerequisites:

• Java Runtime Environment (JDK 8+)
• Maven for building the MCP server
• CData JDBC Driver for Veeva installed and licensed on the host machine
• Access to a Veeva data source (or a configured connection string via the .prp file)

1. Build the MCP server

• Clone the repository and navigate into it git clone https://github.com/cdatasoftware/veeva-mcp-server-by-cdata.git cd veeva-mcp-server-by-cdata

• Build the project with Maven to produce the JAR (includes dependencies) mvn clean install This creates the JAR file: CDataMCP-jar-with-dependencies.jar

2. Install and license the CData JDBC Driver for Veeva

• Download the CData JDBC Driver for Veeva from: https://www.cdata.com/drivers/veeva/download/jdbc
• Install the driver according to the platform instructions
• License the driver:
  • Navigate to the lib folder in the installation directory (examples): Windows: C:\Program Files\CData\CData JDBC Driver for Veeva
    Mac/Linux: /Applications/CData JDBC Driver for Veeva/
  • Run: java -jar cdata.jdbc.veevavault.jar --license
  • Enter your name, email, and a license entry (TRIAL or key)

3. Prepare a .prp configuration file for the JDBC connection

• Create veeva.prp with the required properties, for example: Prefix=veevavault ServerName=CDataVeevaVault ServerVersion=1.0 DriverPath=/path/to/cdata.jdbc.veevavault.jar DriverClass=cdata.jdbc.veevavault.VeevaVaultDriver JdbcUrl=jdbc:veevavault:InitiateOAuth=GETANDREFRESH; Tables=

4. Run the MCP server

• From the project directory, run the JAR with the .prp file: java -jar CDataMCP-jar-with-dependencies.jar /PATH/TO/veeva.prp

5. Connect with an MCP client

• Use Claude Desktop or another MCP client, and add the server entry (as described in the docs).
• Ensure the client locale and file paths match your environment.

Notes:

• The server runs in stdio mode and is intended for local, on-machine clients.
• Ensure the veeva.prp and the JDBC driver JAR have appropriate permissions on the host.

Tips and caveats:

• This is a read-only MCP server by design; the tools expose read operations against the Veeva data source via the JDBC driver.
• The server relies on the configured veeva.prp connection string for JDBC access; ensure the connection string is valid and test it with the CData driver utility before starting the MCP server.
• If Claude Desktop or other clients do not immediately show the MCP server, fully quit and restart the client to reload the MCP configuration.
• The available tools include veeva_get_tables, veeva_get_columns, and veeva_run_query. Outputs are CSV for table/column listings and results, respectively.
• If OAuth is used for the data source, you may need to perform browser-based authentication during the initial connection setup, as described in the setup steps.