Skip to main content

​
Oracle MCP Servers

Reference implementations of Model Context Protocol servers for Oracle products

​
What is MCP?

The Model Context Protocol (MCP) enables standardized, language-agnostic machine-to-machine workflows across data, models, and cloud resources. MCP servers implement specific tool suites, exposing them to MCP-compatible clients like Claude Desktop, Cursor, and Cline. This repository provides 22 production-ready MCP server implementations for Oracle Cloud Infrastructure and database products, demonstrating MCP’s polyglot approach with Python and Java implementations.

​
Key Features

22 MCP Servers

Complete suite of servers for OCI services, MySQL AI, and Oracle Database

Polyglot Architecture

Python and Java implementations demonstrating language flexibility

Dual Transport Modes

Support for both stdio and HTTP streaming transport

OCI Authentication

Integrated authentication with Oracle Cloud Infrastructure

MySQL AI & HeatWave

ML/GenAI tools including vector stores and RAG

Container Ready

Podman/Docker deployment with OAuth2 and TLS support

​
Available Servers

​
Oracle Cloud Infrastructure (OCI)

Connect AI agents to your Oracle Cloud infrastructure with specialized servers for each service:

Generic OCI API

Execute any OCI CLI command through MCP

Compute

Manage compute instances and images

Networking

Configure VCNs, subnets, and security groups

Object Storage

Access and manage object storage buckets

Database

Manage OCI database services

Identity & IAM

Control access and manage users

Monitoring

Access metrics and alarms

Logging

Search and analyze logs

​
Database Servers

MySQL MCP Server

MySQL AI and HeatWave ML/GenAI features including vector stores and RAG

Oracle DB Toolkit

Custom YAML-based tools, JDBC log analysis, and vector similarity search

​
Quick Start

1

Install uv

Install the uv Python package manager:
curl -LsSf https://astral.sh/uv/install.sh | sh
2

Configure OCI Authentication

For OCI servers, set up authentication:
oci session authenticate --region=us-phoenix-1 --tenancy-name=<your-tenancy>
3

Run a Server

Launch any server using uvx:
uvx oracle.oci-api-mcp-server@latest
4

Configure Your MCP Client

Add the server to your MCP client configuration. See Client Configuration for details.

​
Use Cases

Use AI agents to provision compute instances, configure networking, manage storage, and monitor your OCI infrastructure through natural language commands.
Execute SQL queries, perform vector similarity search, use MySQL AI for text generation, and leverage HeatWave ML/GenAI capabilities through conversational interfaces.
Inspect JDBC logs, analyze execution plans, debug database performance issues, and test custom MCP tools using the Oracle Database Toolkit.
Query logs, retrieve metrics, analyze usage patterns, and generate reports across your Oracle Cloud infrastructure using AI-powered assistants.

​
Architecture

Oracle MCP servers follow a polyglot architecture where each server under src/ may be implemented in a different programming language:
  • Python servers: FastMCP-based implementations for OCI and MySQL services
  • Java toolkit: Oracle Database MCP Toolkit with custom YAML-based tool framework
  • Transport modes: Support for both stdio (client-spawned) and HTTP streaming
  • Authentication: Integrated with OCI CLI profiles and session-based auth
These servers are reference implementations intended for exploration, prototyping, and learning. Review the Security Guide before deploying in production environments.

​
Community & Support

Contributing

Learn how to contribute to Oracle MCP servers

Best Practices

Follow our development guidelines and patterns

Security

Review our security policies and disclosure process

License

Released under Universal Permissive License v1.0

Build docs developers (and LLMs) love

Get started for freeTalk to us