# CloudPloy [Health: Active]

**Category:** ☁️ Cloud Platforms  
**Repository:** https://cloudploy.com/mcp  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/cloudploy

## Description
CloudPloy lets your AI agent deploy apps from Claude Code, Cursor, or any MCP client. Add an API key, bring your own Ubuntu or AWS server, or provision Hetzner, DigitalOcean, or AWS at cost. Free forever plan with 1 small server and 1 app.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: medium). Install path inferred — verify against the README before running it. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "cloudploy": {
    "url": "https://cloudploy.com/mcp"
  }
}
```

## Documentation

## What the CloudPloy MCP server does

The CloudPloy MCP server connects an AI agent to CloudPloy’s application deployment workflow. Its primary purpose is to let an agent deploy apps from an MCP-compatible client, including Claude Code and Cursor. The service can work with infrastructure you already provide or with supported cloud resources provisioned through CloudPloy.

For an existing environment, the documented options are an Ubuntu server or an AWS server. CloudPloy can also provision servers from Hetzner, DigitalOcean, or AWS at cost.

## How it works

An MCP client connects to CloudPloy and authenticates with an API key. The agent can then use CloudPloy’s deployment capability to deploy an application. Infrastructure may come from a server supplied by the user or from one of the supported providers.

The product description names Claude Code and Cursor as supported ways to use the CloudPloy MCP server, while also stating that other MCP clients may be used. No transport, endpoint format, tool names, request schema, or authentication variable name is provided, so those details should be confirmed in CloudPloy’s current setup documentation before implementation.

## Setup and configuration

Setup requires an API key. As a result, configuration must be completed using CloudPloy’s own instructions rather than inferred from this listing.

You also need to choose an infrastructure path. Bring an Ubuntu or AWS server if you want to use existing infrastructure, or use CloudPloy to provision Hetzner, DigitalOcean, or AWS resources at cost. The listing does not define server sizes beyond the free-plan reference to one small server.

## Tools and capabilities

The documented capability is agent-driven application deployment. Infrastructure-related options include:

- Using a supplied Ubuntu server
- Using a supplied AWS server
- Provisioning Hetzner infrastructure at cost
- Provisioning DigitalOcean infrastructure at cost
- Provisioning AWS infrastructure at cost

The CloudPloy MCP server is therefore most relevant to developers who want deployment actions exposed through an AI client. The source does not list separate MCP tool names or describe operations such as logs, rollbacks, environment variables, domains, scaling, or database provisioning.

## Limitations and notes

CloudPloy’s free-forever plan includes one small server and one app. Additional infrastructure provisioning is described as being charged at cost. The material does not state whether app limits, server limits, or provider availability differ outside that plan.

Only Claude Code and Cursor are named as clients. Although the description says any MCP client can be used, compatibility details for other clients are not provided. Transport support, operating-system requirements, permissions, regional availability, and security practices are also unspecified in the available source.

_Full upstream README: https://allmcps.com/mcp/cloudploy/readme_

