Oracle Financials Cloud

You run a local, read‑only MCP server that exposes live Oracle Financials Cloud data to LLMs through a simple MCP interface. This enables natural language questions about your cloud data without needing to write SQL, while keeping your data access secure and centralized on your own machine.

How to use

Start the MCP server on your machine and connect your MCP client (for example Claude Desktop) to the local server. The server exposes three core tools you can invoke from your client: a table browser, a column browser, and a query executor. Use these to discover what data is available and to run read‑only queries that surface live Oracle Financials Cloud information through the MCP interface. Typical usage patterns include asking questions like which opportunities are open and which accounts they belong to, or retrieving a list of tables and their columns to understand the data model.

How to install

Prerequisites you need on your machine before starting: a Java Runtime Environment (JRE) compatible with the MCP server and a Java Development Kit (JDK) if you plan to build from source. You will also need a local path to a configured Oracle Financials Cloud data source via the CData JDBC Driver for Oracle Financials Cloud.

Step 1. Build the MCP server locally.

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

Step 2. Install the CData JDBC Driver for Oracle Financials Cloud on your machine following the driver’s installation steps. You will license it once during setup.

Step 3. Create your JDBC connection and store the PRP configuration for the MCP server. The PRP file contains your connection details and driver information.

Step 4. Run the MCP server using the Java command with your PRP file.

java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/oracle-financials-cloud.prp

Additional setup for Claude Desktop (example client)

Configure Claude Desktop to load the MCP server by editing claude_desktop_config.json and adding an entry that points to your local MCP jar and PRP file. The examples below show Windows and Linux/macOS configurations.

{
  "mcpServers": {
    "oraclefin_mcp": {
      "command": "PATH\\TO\\java.exe",
      "args": [
        "-jar",
        "PATH\\TO\\CDataMCP-jar-with-dependencies.jar",
        "PATH\\TO\\oracle-financials-cloud.prp"
      ]
    }
  }
}

{
  "mcpServers": {
    "oraclefin_mcp": {
      "command": "/PATH/TO/java",
      "args": [
        "-jar",
        "/PATH/TO/CDataMCP-jar-with-dependencies.jar",
        "/PATH/TO/oracle-financials-cloud.prp"
      ]
    }
  }
}

Running the server

To run the MCP server directly on your machine, use the following command. The server runs via stdio, so the client and server must be on the same machine.

java -jar /PATH/TO/CDataMCP-jar-with-dependencies.jar /PATH/TO/oracle-financials-cloud.prp

Notes on usage and behavior

Once configured, you can ask questions about the underlying data and the AI client will guide you to read data. You generally do not need to invoke individual tools; simply ask questions, and the client will use the built‑in tools to fetch tables, columns, or run read‑only queries as needed.

The built‑in tools include named endpoints such as get_tables, get_columns, and run_query, which correspond to discovering data structures and extracting data for your questions.

Available tools

oraclefincloud_get_tables

Retrieves a list of tables available in the data source. The output is CSV with headers on the first line.

oraclefincloud_get_columns

Retrieves a list of columns for a given table. The output is CSV with headers on the first line.

oraclefincloud_run_query

Executes a SQL SELECT query against the data source.