DeployHQ

You deploy DeployHQ as an MCP server to let AI assistants interact with your DeployHQ deployments. It provides a secure, typed, and transport-flexible bridge that lets you list projects, servers, and deployments, create new deployments, and fetch logs, all through a clean MCP interface.

How to use

Operate the MCP server from an AI assistant by configuring a local or hosted MCP client to connect via the stdio transport. You will start the MCP server through a package runner and pass credentials through environment variables. Once connected, you can ask your assistant to list projects, inspect servers for a project, view deployment histories, retrieve logs, and create new deployments. The server handles authentication with DeployHQ and provides sane defaults for error handling and logging.

How to install

Prerequisites you need before proceeding:

• Node.js 18+ (Node 20+ recommended)
• Access to a DeployHQ account with API access

Step-by-step installation and start

# Easy install with Claude Code via MCP setup (stdio transport)
claude mcp add --transport stdio deployhq \
  --env DEPLOYHQ_EMAIL=your-email@example.com \
  --env DEPLOYHQ_API_KEY=your-api-key \
  --env DEPLOYHQ_ACCOUNT=your-account -- npx -y deployhq-mcp-server

This configuration uses a stdio transport. The server runs via npx and the environment variables provide the necessary credentials without storing them.

Configuration

Use the following config to run the MCP server locally or in your Claude Code project. It specifies the stdio runtime and the required credentials.

{
  "mcpServers": {
    "deployhq": {
      "command": "npx",
      "args": ["-y", "deployhq-mcp-server"],
      "env": {
        "DEPLOYHQ_EMAIL": "your-email@example.com",
        "DEPLOYHQ_API_KEY": "your-api-key",
        "DEPLOYHQ_ACCOUNT": "your-account-name",
        "LOG_LEVEL": "INFO"  
      }
    }
  }
}

Security and operational notes

Credentials are supplied via environment variables and are not stored by the MCP server. Use least-privilege API keys and consider restricting read/write access as needed. Enable verbose logging only when troubleshooting by setting LOG_LEVEL to DEBUG.

Troubleshooting

If the server fails to start, verify that required environment variables are set and that Node.js 18+ is installed. Look for error messages in your Claude client logs and enable DEBUG-level logs if needed by setting LOG_LEVEL=DEBUG.

Common usage scenarios

With the server connected, you can instruct your assistant to perform typical tasks such as listing all DeployHQ projects, showing servers for a specific project, retrieving the latest deployment status, creating a new deployment, or retrieving deployment logs for debugging.

Notes on hosted deployment (optional)

If you host the MCP server with a web transport, you can expose an HTTP or SSE interface for integration. Follow standard hosting setup for your platform and ensure your environment variables remain protected.

Existing commands you’ll use during setup

When you configure Claude Code, you’ll typically supply the stdio runtime and credentials in the client configuration, as shown in the example above. The actual runtime command used by the MCP server in this setup is npx with the deployhq-mcp-server package.

Notes on environment variables

Required: DEPLOYHQ_EMAIL, DEPLOYHQ_API_KEY, DEPLOYHQ_ACCOUNT. Optional: LOG_LEVEL, NODE_ENV. These variables control authentication, log verbosity, and runtime behavior.

What you get with the DeployHQ MCP Server

A typed, secure bridge to the DeployHQ REST API that supports multiple transports, easy setup with MCP clients, and production-ready error handling and logging.

Available tools

list_projects

List all projects in your DeployHQ account. Returns an array with repository information and deployment status.

get_project

Get detailed information about a specific project using its permalink or identifier.

list_servers

List all servers configured for a given project.

list_deployments

List deployments for a project with pagination support.

get_deployment

Get detailed information about a specific deployment.

get_deployment_log

Get the deployment log output for debugging failed deployments.

create_deployment

Create a new deployment for a project with configurable parameters.