oracle/mcp
Repository containing MCP (Model Context Protocol) servers that provides a suite of tools for managing and interacting with Oracle products.
What it solves
This project provides reference implementations of Model Context Protocol (MCP) servers that allow AI models to interact with Oracle products and Oracle Cloud Infrastructure (OCI). It solves the problem of providing a standardized, language-agnostic way for AI clients to access and manage cloud resources and data without needing custom integrations for every tool.
How it works
The repository contains a polyglot architecture where different MCP servers are implemented in various languages (Python, Node.js, Java). These servers act as intermediaries that expose specific tool suites to MCP-compatible clients (like Cursor, Cline, or MCPHost). They can connect to Oracle services via the official OCI Python SDK or the OCI CLI, supporting both stdio and HTTP streaming transport modes.
Who it’s for
- AI Developers building agents that need to manage Oracle Cloud resources.
- Developers exploring the Model Context Protocol for prototyping and learning.
- Users of MCP-compatible IDEs (such as Cursor or Cline) who want to integrate Oracle Cloud workflows into their AI-assisted coding environment.
Highlights
- Polyglot Support: Implements servers in Python, Node.js, and Java to demonstrate the protocol's language-agnostic nature.
- Multiple Transport Modes: Supports both
stdiofor local execution and HTTP streaming for remote connectivity. - OCI Integration: Provides specialized servers for general OCI workflows, OCI CLI command discovery, and specific Oracle product domains.
- Containerized Deployment: Includes support for running servers via Podman for easier environment isolation.
Related
- Project
- Project
- Project
- Project
- Project