Railway MCP Server

A Model Context Protocol (MCP) server for integrating with the Railway.app platform.

Table of Contents

Features

• ✅ Authentication with Railway API tokens
• ✅ Project management (list, info, delete)
• ✅ Deployment management (list, restart)
• ✅ Service management (create from GitHub repo or Docker image, list)
• ✅ Variable management (list, create/update, delete)
• ✅ Service Network management
• ✅ Volume management
• ❌ Full support for all templates
  • 🚧🔨⏳ Database template support
  • Automatic database and networking workflows
• 🚧🔨⏳ Most commonly used workflows
• ❌ Automatic GitHub repository linking for services

Installation

Prerequisites

• Node.js 18+ (for built-in fetch API support)
• An active Railway account
• A Railway API token (create one at https://railway.app/account/tokens)

Quick Start

This MCP server is designed to work with MCP Clients like:

• Claude for Desktop | ✅ Battle-Tested
• Cursor | ✅ Needs Testing
• Cline | 🚧🔨⏳ Needs Testing
• Windsurf | 🚧🔨⏳ Needs Testing
• Other MCP Clients | 🚧🔨⏳ Needs Testing

Installing via Smithery

To install railway-mcp automatically, we recommend using Smithery

Claude Desktop

Cursor

1. Head to your cursor settings and find the MCP section
2. Click 'Add new MCP server'
3. Name it however, you like, we recommend railway-mcp for better clarity
4. Paste this command into the 'Command' section, where is your accounts Railway token:

1. Create or edit your Claude for Desktop config file:
   • macOS: ~/Library/Application\ Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json
2. Add the railway-mcp server to your configuration with your API token:

When you have multiple MCP servers, your config file might look like this:

3. Restart Claude for Desktop
4. You can now start using Railway tools directly in Claude. For example:

5. Alternatively, if you don't want to add your token to the configuration file, you can configure it within Claude using:

Recommendations and Other Information

This server best combines with MCP-clients that have access to terminal or with Git (Cursor, Windsurf). Using this MCP with others is recommended as railway-mcp orchestrates containers and streamlines your deployment process seamlessly.

Recommended MCP servers to combine with

• Git || Official Link
• GitHub || Official || Smithery

For Claude

• Out of the box, Claude does not have terminal access, so it cannot trigger deployments as it will not be able to get the latest commit.
• Spinning up different services and monitoring them are the best use case with Claude.

For Cursor

• Use with GitHub MCP or have the repository already setup on GitHub and cloned locally on your machine to leverage full integration with railway-mcp.
• When Cursor makes a change, it may forget to push it's changes to GitHub causing it to try and deploy a commit that Railway cannot pull.
  • SOLUTION: Always ask or include somewhere in your prompt: Have you pushed our changes to GitHub yet?

Security Considerations

• Railway API tokens provide full access to your account. Keep them secure.
• When using the environment variable method, your token is stored in the Claude Desktop configuration file.
• Sensitive variable values are automatically masked when displayed.
• All API calls use HTTPS for secure communication.
• The server's memory-only token storage means your token is never written to disk outside of the configuration file.

Troubleshooting

If you encounter issues:

1. Token Authentication Issues
   • Ensure your API token is valid and has the necessary permissions
   • If using the environment variable method, check that the token is correctly formatted in the config file
   • Try using the configure tool directly in Claude if the environment token isn't working
2. Server Connection Issues
   • Check that you've installed the latest version of the server
   • Verify that Node.js version 18 or higher is installed
   • Restart Claude for Desktop after making changes to the configuration
3. API Errors
   • Verify that you're using correct project, environment, and service IDs
   • Check Railway's status page for any service disruptions
   • Railway API has rate limits - avoid making too many requests in a short period

Contributing

We welcome contributions from the community! Please see our Contributing Guidelines for details on how to get started, development guidelines, and debugging information.

Available Tools

Authentication

• configure - Set your Railway API token (only needed if not provided in environment variables)

Projects

• project-list - List all projects in your account
• project-info - Get detailed information about a specific project
• project-create - Create a new project with optional team ID
• project-delete - Delete a project
• project-environments - List all environments in a project

Services

• service-list - List all services in a specific project
• service-info - Get detailed information about a specific service
• service-create-from-repo - Create a new service from a GitHub repository
• service-create-from-image - Create a new service from a Docker image
• service-delete - Delete a service from a project
• service-restart - Restart a service in a specific environment
• service-update - Update service configuration (build command, start command, etc.) | 🚧 Needs Testing

Deployments

• deployment-list - List recent deployments for a service
• deployment-trigger - Trigger a new deployment for a service
• deployment-logs - Get logs for a specific deployment
• deployment-health-check - Check the health/status of a deployment

Variables

• variable-list - List variables for a service or environment
• variable-set - Create or update a variable
• variable-delete - Delete a variable
• variable-bulk-set - Bulk update variables for a service | 🚧 Needs Testing
• variable-copy - Copy variables between environments | 🚧 Needs Testing

Databases

• database-list-types - List all available database types that can be deployed
• database-deploy - Deploy a new database service

Setting up a new service

1. List projects to get the project ID
2. Create a new service from a template
3. Add environment variables
4. View the service deployment

Managing environment variables

1. List projects to find your project ID
2. List variables to see what's currently set
3. Create or update variables as needed
4. Delete any obsolete variables