Get the FREE Ultimate OpenClaw Setup Guide →

mcp

Repository containing MCP (Model Context Protocol) servers that provides a suite of tools for managing and interacting with Oracle products.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio oracle-mcp uvx oracle.mcp-server@latest \
  --env FASTMCP_LOG_LEVEL="INFO"

How to use

This repository contains reference MCP (Model Context Protocol) servers for Oracle products and other language implementations. Each server under src/ is a distinct MCP server that demonstrates language-agnostic MCP concepts and provides a working example of how clients can interact with tool suites via MCP. To use a specific server, choose a server directory under src/ that matches your preferred language (e.g., Python, Node.js, Java) and follow its individual README for setup and runtime details. The server exposes MCP endpoints that can be consumed by MCP clients, and the repository includes example client configurations and instructions for connecting over HTTP/streaming transports.

Typical workflow:

  • Pick a server under src/ and read its README for prerequisites and runtime requirements.
  • Run the server using the recommended command (often via uvx for Python or the project’s package manager).
  • Configure your MCP client with the server name, transport type, and connection URL as demonstrated in the examples. The client config is usually placed in a JSON file with an mcpServers object describing each server’s command, args, and environment variables when necessary.

How to install

Prerequisites:

  • A supported OS (Linux, macOS, or Windows depending on the server)
  • Git
  • Internet access
  • Language runtimes and package managers required by the chosen server (e.g., Python, Node.js, Java, etc.)

Steps:

  1. Clone the repository: git clone https://github.com/your-org/oracle-mcp.git cd oracle-mcp

  2. Review the README for the specific server you want to run (located under src/<server-name>/README.md) to confirm language/runtime requirements and any OCI authentication steps if applicable.

  3. Install the prerequisites for the chosen server (as described in its README). For example, if using a Python-based server, install uv and the Python runtime; for a Node.js server, install Node.js and npm/yarn as required.

  4. Start the server using the recommended command from the server’s README. Common patterns include:

  • Python/uv-based server: uvx <server-name>@latest
  • Node.js server: node path/to/server.js
  • Docker/podman: follow the containerization instructions in the server’s README if provided
  1. Configure your MCP client with the appropriate mcpServers entry (see the Quick Start examples in the root README) to connect to the running server.

Additional notes

Notes and tips:

  • This repository provides reference implementations; check each server’s README for exact prerequisites and version requirements.
  • If you intend to connect via OCI-related servers, you may need OCI CLI credentials and configuration as outlined in the Authentication section of the root README.
  • Transport type and URL may differ between clients; some clients expect http transport, others streamableHttp. Use the example configurations provided for your client.
  • When mounting or sharing credentials (e.g., OCI config), use environment variables or proper path mappings as shown in the examples. Use ~ in paths inside containers to ensure cross-environment path expansion.
  • For podman/docker-based runs, ensure proper volume mounts and environment variable propagation to the container as demonstrated in the repository’s podman sections.

Related MCP Servers

Sponsor this space

Reach thousands of developers