CLI

The Salus CLI is a single binary, salus. Its headline feature is exposing Salus to your AI coding tools as an MCP server, so assistants like Claude Code, Cursor, Codex, OpenCode, and Claude Desktop can work with your projects, deployments, logs, databases, and pipelines on your behalf.

Install

Install the salus binary with the one-line installer:

curl -fsSL https://download.salus.cloud/install.sh | bash

This installs salus to /usr/local/bin/salus. Prefer not to use the terminal? Download the Claude Desktop extension instead — the .mcpb bundle at https://download.salus.cloud/salus-mcp.mcpb — and install it into Claude Desktop.

Connect it to your AI tools

Run the CLI as an MCP server with the mcp subcommand. Point your editor at the salus binary and it exposes Salus as a set of tools. For Claude Code, add this to .mcp.json:

{
  "mcpServers": {
    "salus": {
      "command": "/usr/local/bin/salus",
      "args": ["mcp"]
    }
  }
}

Configuration snippets for Cursor, Codex, and OpenCode are available on the Download page — each runs the same salus mcp command.

Sign in

There’s no separate login step. The first time your assistant uses Salus, the MCP server prompts you to sign in — it opens your browser to complete authentication, and gives you a link if the browser can’t open automatically. Your session is reused after that.

What your assistant can do

Once connected, your assistant can act on the platform concepts you already know: projects, deployments, logs, and databases.

You choose how much access it has, from Read-only through Standard, Operator, and Full — or a custom selection. Start restrictive and widen access only as you need it, especially for tools that can change or deploy your applications.