CloudPloy is a deployment automation tool that enables users to deploy applications through AI agents. The platform integrates with AI coding tools such as Claude Code, Cursor, and any MCP (Model Context Protocol) client, allowing developers to initiate deployments directly from their AI assistant. Users create an API key through the CloudPloy dashboard, grant deploy and provision permissions, and then point their AI tool at the CloudPloy service to begin deploying applications to their chosen infrastructure. The platform supports two infrastructure approaches: users can SSH into a server they already own, or they can provision new compute resources through AWS. CloudPloy offers a flat pricing plan for its control plane, with compute resources provided at cost. The tool includes a set of server management capabilities, including listing servers, estimating costs, viewing account details, provisioning new servers, and stopping servers. These operations are accessible through the same MCP interface or through a REST API at /v1. The typical workflow involves three steps: creating an API key with appropriate permissions in the team settings, configuring the AI tool to connect to CloudPloy, and then deploying to the user's cloud infrastructure. The platform is designed to make the AI agent the primary path for deployment, with an optional web panel available for additional management. CloudPloy is currently operating on a waitlist basis, with access granted as spots open.