Install Severalnines CCX

Pick your client, copy the command, done.

claude mcp add -e "CCX_BASE_URL=${CCX_BASE_URL}" ccx-mcp -- npx -y @severalnines/ccx-mcp

Required:CCX_BASE_URL+ 4 optional

Environment Variables

Set these before running Severalnines CCX.

VariableDescriptionRequired

CCX_BASE_URLThe URL of your CCX deploymentYes

CCX_USERNAMEYour CCX account emailNo

CCX_PASSWORDYour CCX account passwordNo

CCX_CLIENT_IDOAuth2 client IDNo

CCX_CLIENT_SECRETOAuth2 client secretNo

Once configured, Severalnines CCX gives your AI agent access to:

ccx_list_datastoresList all database clusters with status, vendor, and cloud provider

ccx_get_datastoreGet detailed cluster info including credentials and job progress

ccx_create_datastoreCreate a new cluster

ccx_delete_datastoreDelete a cluster

ccx_get_nodesGet cluster nodes with roles, status, and IP addresses

ccx_get_connection_stringConnection strings in URI, CLI, JDBC, and env formats

ccx_scale_datastoreScale instance size or expand storage volume

ccx_add_nodeAdd a new replica node to a cluster

ccx_list_cloudsList available cloud providers and regions

ccx_list_plansList instance sizes and volume types

After setup, try these prompts with your AI agent:

→List my datastores

→Create a PostgreSQL cluster on AWS in eu-west-1

→Get the connection string for my production database

→Scale my cluster to a medium instance

→What's the CPU usage on my database?

Prerequisites & system requirements

An MCP-compatible client (Claude Code, Cursor, Windsurf, Claude Desktop, or Codex)
Node.js 18+ with npm/npx installed
CCX_BASE_URL — The URL of your CCX deployment

Alternative installation methods

npm install @severalnines/ccx-mcp

git clone https://github.com/severalnines/ccx-mcp.git && cd ccx-mcp && npm install && npm run build

Conare · memory for coding agents

Save the docs, env vars, and workflow around Severalnines CCX in Conare so Claude Code, Codex, and Cursor remember it next time.