Add the Oracle Analytics Cloud MCP Server to Your AI Client

This topic explains how to connect your AI client, for example Claude, Cline, or Copilot to the Oracle Analytics Cloud MCP server.

  1. Confirm that the MCP connection utility is installed on your computer. See Download and Install the MCP Connection Utility.
  2. Install the required host AI client and any required extensions on your computer. For example, install Cline and install the Visual Studio Code extension.
  3. Go back to your Oracle Analytics Cloud home page, click the name badge with your initials, and click Profile.
  4. In the Profile page click the MCP Connect tab and copy the server configuration:
    • If your MCP server configuration file doesn't contain any server configurations, click the Copy JSON button to copy the required JSON object.
    • If your MCP server configuration file already contains server configurations, manually copy only the 'mcpServers' property.
  5. In your host AI client, navigate to and open the MCP server configuration file, for example claude_desktop_config.json.
  6. Paste the copied MCP server configuration JSON into the configuration file. Make sure to add the new entry carefully without modifying or overwriting any existing properties.
  7. Replace "path/to/oac-mcp-connect.js" with the location of the oac-mcp-connect.js file on your computer. For example, "/Users/user_name/oac-mcp-connect/oac-mcp-connect.js".
  8. Save the configuration file and restart the AI client.
  9. In the AI client:
    • Confirm that oac-mcp-server is included and enabled in the client's list of MCP servers.
    • Confirm that the Oracle Analytics Cloud Discover, Describe, and Execute tools are included and enabled.