Dokploy/mcp

The Official MCP package for Dokploy

What it solves

Dokploy MCP Server allows AI models and MCP-compatible clients (like Cursor, VS Code, and Claude Desktop) to programmatically interact with a Dokploy server. It bridges the gap between LLMs and your infrastructure management, enabling AI to manage projects, applications, and databases via the Dokploy API.

How it works

The server implements the Model Context Protocol (MCP), exposing 508 tools across 49 categories that map directly to Dokploy API endpoints. It can be deployed via Node.js (npx), Docker, or Bun/Deno, and supports two transport modes:

  • Stdio Mode: Direct process communication for desktop apps.
  • HTTP Mode: Streamable HTTP and legacy SSE for web applications.

It uses TypeScript and Zod for type-safe schema validation and allows users to filter which tools are loaded using the DOKPLOY_ENABLED_TAGS environment variable.

Who it’s for

Developers and DevOps engineers who use Dokploy for server management and want to automate infrastructure tasks using AI agents or LLM-integrated IDEs.

Highlights

  • Comprehensive API Coverage: Provides 508 tools covering everything from application lifecycles and Docker Compose to database management (Postgres, MySQL, MongoDB, etc.).
  • Flexible Deployment: Supports npx, Docker, Bun, Deno, and multiple MCP clients (Cursor, Windsurf, VS Code, Claude Code, Zed, BoltAI).
  • Security Features: Includes a redaction system (DOKPLOY_REDACT_ENV) to prevent sensitive environment variables and secrets from being sent to the LLM.
  • Semantic Annotations: Tools include hints like readOnlyHint and destructiveHint to help AI models understand the safety and impact of an action.