MCP Server

The Simplifyd Cloud MCP Server is a Model Context Protocol (MCP) server that enables natural language interaction with your Simplifyd Cloud workspaces, projects, services, and deployments.

With it, you can ask your AI assistant to deploy a service, manage environment variables, scale a container, check deployment logs, and much more — without leaving your editor.

The MCP server is open-source and lives in the mcp/ directory of the cloud monorepo.

Understanding MCP and the Simplifyd Cloud MCP Server

The Model Context Protocol (MCP) defines a standard for how AI applications (hosts) can interact with external tools and data sources through a client–server architecture.

• Hosts: Applications such as Claude Code, Cursor, VS Code, or Claude Desktop that connect to MCP servers.
• Clients: The layer within hosts that maintains one-to-one connections with individual MCP servers.
• Servers: Standalone programs (like the Simplifyd Cloud MCP Server) that expose tools and workflows to AI assistants.

The Simplifyd Cloud MCP Server acts as the server in this architecture, translating natural language requests from your AI assistant into authenticated REST API calls against the Simplifyd Cloud API.

Prerequisites

• Go 1.23+ to build the server binary, or a pre-built binary from the releases page
• A Simplifyd Cloud account — sign up at cloud.simplifyd.com
• A valid API token (or your account credentials to authenticate interactively)

Build

Clone the monorepo and build the binary:

git clone https://github.com/simplifyd-com/cloud
cd cloud/mcp
go build -o cloud-mcp .

Note the full path to the cloud-mcp binary — you will need it when configuring your MCP client below.

Installation

claude mcp add simplifyd-cloud /path/to/cloud/mcp/cloud-mcp \
  -e SIMPLIFYD_API_TOKEN=<your-token>

claude mcp list

{
  "mcpServers": {
    "simplifyd-cloud": {
      "command": "/path/to/cloud/mcp/cloud-mcp",
      "env": {
        "SIMPLIFYD_API_TOKEN": "<your-token>"
      }
    }
  }
}

{
  "servers": {
    "simplifyd-cloud": {
      "type": "stdio",
      "command": "/path/to/cloud/mcp/cloud-mcp",
      "env": {
        "SIMPLIFYD_API_TOKEN": "<your-token>"
      }
    }
  }
}

{
  "mcpServers": {
    "simplifyd-cloud": {
      "command": "/path/to/cloud/mcp/cloud-mcp",
      "env": {
        "SIMPLIFYD_API_TOKEN": "<your-token>"
      }
    }
  }
}

Obtaining a token

You can get your API token from any of these sources:

• Login tool — omit SIMPLIFYD_API_TOKEN from the config and use the login MCP tool on first use. The server will save the token to ~/.simplifyd/config.json and reuse it automatically.
• Dashboard — go to Project Settings → Tokens, create a token scoped to your environment, and copy it.
• CLI — run edge token create --name "MCP" and copy the output.

Example usage

Once the server is connected, you can interact with your infrastructure in plain English. Below are some examples.

Deploy a Docker service:

"Create a service called api using the nginx:latest image in my production environment and deploy it."

Check deployment status:

"What's the status of the latest deployment for the api service?"

Tail deployment logs:

"Show me the logs for the last deployment of api in the staging environment."

Manage environment variables:

"Set DATABASE_URL and REDIS_URL on the api service in production."

Scale a service:

"Scale the api service to 3 replicas and give it 2 vCPUs."

Add a public URL:

"Add an HTTPS ingress on port 8080 for the api service."

Invite a team member:

"Add alice@example.com to my acme workspace."

Check billing:

"How much have I spent this month in the acme workspace?"

Available MCP tools

The server exposes 45 tools grouped by resource type.

Authentication

Workspaces

Projects

Environments

Services

Deployments

API Tokens

Security considerations

• Review AI actions before they execute. Most MCP clients show you the tool call and its arguments before running. Read them carefully, especially for destructive operations like delete-service or undeploy-service.
• Use scoped tokens. Create a dedicated token with access to only the environments the AI assistant needs. Avoid using your personal JWT from edge auth login.
• Restrict to non-production environments when experimenting. Give the AI access to a staging or dev environment first until you are confident in its behaviour.
• Do not share your ~/.simplifyd/config.json. It contains your authentication token in plain text.
• Rotate tokens regularly. If you suspect a token has been compromised, revoke it immediately from the dashboard or with edge token delete <slug>.

Feature requests

The MCP server is open source. If a tool or workflow is missing, open an issue or pull request in the cloud monorepo.