Skip to content
Raindrop Developer Hub

Raindrop MCP Quick Start

The Raindrop MCP server transforms AI coding tools into powerful infrastructure development platforms. Instead of just writing code, you can build complete applications with databases, APIs, and deployment infrastructure through structured, guided workflows.

Step 1: Install AI Tool

Choose your preferred AI coding tool and follow the installation instructions:

Install Claude Code

Terminal window
npm install -g @anthropic-ai/claude-code

Step 2: Install Raindrop CLI

Terminal window
npm install -g @liquidmetal-ai/raindrop

Step 3: Sign Up for Raindrop

Create your Raindrop account:

  1. Sign up for a Raindrop account at liquidmetal.run
  2. Complete the account setup process

Step 4: Authenticate Raindrop CLI

Run the following command to authenticate the Raindrop CLI:

Terminal window
raindrop auth login

Step 5: Set Up MCP Integration

Run the installation command for your AI tool:

Terminal window
raindrop mcp install-claude

What this command does:

  • Configures Claude Code with the correct MCP server URL
  • Sets up raindrop.md and custom slash commands for Claude Code (/new-raindrop-app, /update-raindrop-app, and /reattach-raindrop-session)

Step 6: Start Your AI Tool

  1. Start Claude Code

  2. Run the MCP command:

/mcp

  1. Select the raindrop-mcp server

  2. Select Authenticate

  3. Login to your Raindrop account (or click Allow Access if you are already logged in)

Using the Raindrop MCP Server

Once configured, you have access to building complete applications through a guided workflow.

Start a New Application

Create a new Raindrop application from scratch:

Using Slash Command:

/new-raindrop-app

Using Direct MCP Call:

Claude, call the LiquidMetal Raindrop:login MCP tool to start a new application development session.

What Happens:

  • Creates a new session ID and timeline ID
  • Guides you through requirements gathering (PRD creation)
  • Generates application architecture and manifest
  • Sets up databases, services, and infrastructure
  • Implements complete application code
  • Deploys to live infrastructure
  • Validates deployment with endpoint testing

Next Steps