StuMason/coolify-mcp

MCP server for Coolify — 42 optimized tools for managing self-hosted PaaS through AI assistants

What it solves

It allows users to manage their Coolify self-hosted PaaS (Platform-as-a-Service) using natural language through AI clients like Claude, Cursor, or any MCP-compatible tool. It bridges the gap between a complex infrastructure management interface and an AI agent, enabling deployment, debugging, and operation of servers and applications without manual API calls or dashboard navigation.

How it works

The project implements a Model Context Protocol (MCP) server that exposes 46 specialized tools to an AI model. These tools interface with the Coolify API to perform actions such as deploying apps, reading logs, and managing environment variables. It supports three deployment modes: a one-click installation for Claude Desktop, a local installation via npx, and a remote containerized mode using HTTP/OAuth 2.1. It also includes a doctor command to verify configuration and connectivity.

Who it’s for

  • DevOps engineers and developers who want to manage their infrastructure via chat.
  • Agencies managing multiple Coolify instances for different clients (via "fleet mode").
  • Users of self-hosted PaaS who prefer an agentic workflow for troubleshooting and deployment over a manual CLI or GUI.

Highlights

  • Comprehensive Toolset: 46 tools for diagnosing apps, deploying/rolling back, creating resources, and handling configurations.
  • Safety First: Destructive operations require human confirmation in supported clients, and secrets are masked at the API boundary.
  • Multi-Instance Support: Ability to manage a "fleet" of multiple Coolify instances from a single connection.
  • Built-in Workflows: Includes pre-defined prompts for troubleshooting applications, explaining failed deployments, and checking estate health.
  • Connectivity Diagnostics: A doctor tool that identifies common configuration errors and provides one-line fixes.

Related

  • Project
    Dokploy/mcp

    An MCP server that exposes the full Dokploy API as tools for AI models, enabling programmatic infrastructure and application management.

  • Project
    bvisible/mcp-ssh-manager

    An MCP server that enables AI agents to securely manage remote servers via SSH, featuring per-server security modes to control command execution.

  • Project
    googleapis/mcp-toolbox

    MCP Toolbox is an open‑source server that lets AI agents and IDEs talk to enterprise databases via the Model Context Protocol. It offers ready‑made database tools and a framework for defining custom, secure SQL/semantic‑search tools, with SDKs for Python, JS/TS, Go, and Java.

  • Project
    cloudflare/skills

    A collection of agent skills and MCP servers that equip AI coding assistants with specialized knowledge of the Cloudflare Developer Platform for better guidance and deployment.

  • Project
    devopness/devopness

    A cloud infrastructure and deployment platform that allows users to provision servers and deploy apps across multiple clouds, featuring an MCP server for AI-agent managed DevOps.