home / mcp / aiven mcp server

Aiven MCP Server

Provides access to Aiven services and enables actions to list projects, list services, and get service details.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "ampcome-mcps-aiven-mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "$REPOSITORY_DIRECTORY",
        "run",
        "--with-editable",
        "$REPOSITORY_DIRECTORY",
        "--python",
        "3.13",
        "mcp-aiven"
      ],
      "env": {
        "AIVEN_TOKEN": "$AIVEN_TOKEN",
        "AIVEN_BASE_URL": "https://api.aiven.io"
      }
    }
  }
}

This MCP Server lets you access Aiven services such as PostgreSQL, Kafka, ClickHouse, Valkey, and OpenSearch from your AI workflows. It exposes practical tools to list projects, enumerate services, and fetch details, enabling you to build end-to-end solutions that interact with your Aiven ecosystem.

How to use

You interact with the Aiven MCP Server from an MCP client. Use the provided tools to discover your Aiven setup and obtain service details, then integrate those results into your workflows. The available actions focus on listing projects, listing services within a project, and retrieving detailed information about a specific service.

How to install

Prerequisites you need before installing include Python 3.13 and the MCP runtime environment. You’ll also need access to the repository that contains the MCP server and the ability to run the MCP runtime (uv).

Step-by-step commands you should run in your environment:

# 1. Set up your environment (adjust paths as needed)
# Ensure Python 3.13 is available
python3.13 --version

# 2. Obtain the MCP server code
# Clone the repository that contains the MCP server (replace with your actual URL or path)
# git clone <REPOSITORY_URL>
# cd ampcome-mcps/aiven-mcp

# 3. Install and synchronize dependencies using uv
uv sync

# 4. Activate the local Python virtual environment (if applicable)
# source .venv/bin/activate

# 5. Start the MCP server for testing (development path)
mcp dev mcp_aiven/mcp_server.py

Additional configuration and security notes

Environment variables used to connect to Aiven must be configured in your environment. The MCP server relies on two required values: AIVEN_BASE_URL for the Aiven API URL and AIVEN_TOKEN for authentication. Treat these credentials with care and follow your organization’s secret-management policies.

Tools exposed by the MCP server

The server provides the following actions to interact with Aiven programmatically:

Security and access considerations

MCPs run in your environment, so you are responsible for their security and compliance. The AI agents you deploy will have permissions limited by the API token you provide. Always apply the principle of least privilege to tokens and carefully manage credentials that may be shared with AI agents.

Notes on configuration for local testing

To run this MCP server locally, you need to configure an MCP client to load the standard MCP entry for the Aiven server. The common setup uses a local uv command that points to the repository and starts the MCP server as shown in the examples.

Available tools

list_projects

List all projects on your Aiven account.

list_services

List all services in a specific Aiven project.

get_service_details

Get the detail of your service in a specific Aiven project.