selvpc (Selectel VPC)
python-selvpcclient
Python client and selvpc CLI for the Selectel cloud platform API
Selectel: skills, MCP and CLI
There is an official tool from Selectel called selvpc (Selectel VPC) that you should install first for working with Selectel from an AI agent. It provides a Python client and CLI for the Selectel cloud platform API. Besides that, there are several MCP servers for different tasks: selectel-mcp for OpenStack, S3, billing, and Docker deploys; Selectel balance MCP server for billing and spend predictions; and Selectel Cloud Platform MCP server for managing projects, resources, and quotas. These MCPs support agents like Claude Code, Cursor, SourceCraft, Koda, and Coddy.
Checked on
Official
These tools come from the service vendor itself.
MCP servers
An unofficial Selectel cloud MCP server: OpenStack servers, S3 storage, billing and one-call Docker app deploys
Selectel balance MCP server
Selectel MCP Server
A Selectel billing MCP server: balance and spend predictions via a service user, with automatic token refresh and an audit log
Selectel Cloud Platform MCP server
Selectel MCP Servers
A FastMCP Selectel Cloud Platform server for projects, resources and quotas, planned as the first of six service-specific MCPs
Frequently asked questions
Is there an official MCP for Selectel?
No official MCP server from Selectel is listed, only the official CLI tool selvpc. The MCP servers are unofficial community tools.
How to connect Selectel to Claude Code or Cursor?
Use selectel-mcp or Selectel balance MCP server for Claude Code, and Selectel Cloud Platform MCP server for Cursor. These MCPs provide API access tailored for these agents.
Which MCP for Selectel should I choose?
Choose selectel-mcp for OpenStack and storage tasks with Claude Code, or Selectel balance MCP server for billing with Claude Code. For project and quota management with Cursor, use Selectel Cloud Platform MCP server.
Is it safe to give an AI agent access to Selectel?
The official selvpc tool has a high risk level. The MCP servers have medium to high risk, so caution is advised when granting access to AI agents.