RekklesNA/ProxmoxMCP-Plus

What it solves

ProxmoxMCP-Plus provides a unified, security-conscious control plane for managing Proxmox VE environments. It eliminates the need for operators to manually combine raw API calls, shell scripts, and custom polling for common operational workflows like VM and LXC management, snapshots, and backups.

How it works

The project acts as an intermediary layer between AI clients and the Proxmox VE API. It exposes its functionality through two primary access paths:

• MCP (Model Context Protocol): Enables AI agents (such as Claude Desktop, Cursor, and VS Code) to interact with Proxmox using natural language.
• OpenAPI: Provides a standard HTTP interface for automation, dashboards, and no-code tools.

It includes a persistent job store (using SQLite) to track long-running asynchronous Proxmox tasks via stable job_ids, allowing clients to poll, retry, or cancel operations.

Who it’s for

• AI Agent Operators: Users who want to control their virtualization infrastructure via LLM-powered agents.
• DevOps/Sysadmins: Professionals managing Proxmox VE who need a standardized OpenAPI bridge for automation.
• Home Lab Enthusiasts: Users seeking a simplified way to manage VMs and containers with integrated job tracking.

Highlights

• Dual Access Paths: Native MCP for AI agents and OpenAPI for standard HTTP automation.
• Comprehensive VM/LXC Lifecycle: Support for creation, starting, stopping, deleting, and snapshotting of VMs and LXCs.
• LXC Command Execution: SSH-backed command execution for containers with integrated policy controls.
• Job Persistence: Stable tracking of asynchronous tasks with audit history and retry capabilities.
• Security-First Design: Support for Proxmox API tokens, Bearer authentication, TLS validation, and tool allowlists/denylists.