home / mcp / oracle mcp server

Oracle MCP Server

mcp服务器oracle数据库连接

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "anpy-j-mcp-oracle": {
      "command": "uvx",
      "args": [
        "mcp-server-oracle"
      ],
      "env": {
        "ORACLE_CONNECTION_STRING": "username/password@hostname:service_name"
      }
    }
  }
}

You can run the MCP server for Oracle model access locally and connect it to your MCP client to retrieve data from an Oracle database. This server lets you expose Oracle data and run queries as part of your broader MCP workflow, enabling seamless integration with your AI-assisted tools.

How to use

You use this MCP server by starting it through your preferred MCP client setup and pointing the client to the local stdio command. The server expects a single environment variable that provides the Oracle connection string. Once running, your MCP client can query Oracle data and perform actions that the server supports through the MCP protocol.

How to install

Prerequisites include UV as your package manager, Python 3.12 or later, and an MCP client environment such as Claude Desktop.

1. Install the required package manager and dependencies for your environment.

2. Configure your MCP client to include the Oracle MCP server as a stdio endpoint using the provided command and arguments.

Configuration and startup

Add the following server configuration to your MCP client config to run the Oracle MCP server locally via stdio.

{
  "mcpServers": {
    "oracle": {
      "command": "uvx",
      "args": [
        "mcp-server-oracle"
      ],
      "env": {
        "ORACLE_CONNECTION_STRING": "username/password@hostname:service_name"
      }
    }
  }
}