golive-mcp
Allows deploying to Cloudflare Pages projects.
Allows deploying to cPanel hosting environments.
Allows deploying to DigitalOcean App Platform and Droplets.
Allows deploying to Fly.io applications.
Allows deploying to GitHub Pages sites.
Allows deploying to Google Cloud Run and Google Compute Engine.
Allows deploying to Heroku applications.
Allows deploying to Hetzner VPS instances.
Allows deploying to Netlify projects.
Allows deploying to OVH VPS instances.
Allows deploying to Plesk hosting panels.
Allows deploying to Railway projects.
Allows deploying to Render services.
Allows deploying to Spaceship (cPanel shared hosting).
Allows deploying to Vercel projects.
Allows deploying to Vultr VPS instances.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@golive-mcpdeploy to production"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
golive-mcp
Go Live — deploy from your AI agent in one command.
Works with Claude Code, Cursor, Gemini CLI, Copilot CLI, or any MCP-compatible client.
What is this?
Go Live is an MCP server that gives your AI agent superpowers: create a GitHub PR, deploy to production, and verify it's live — all from a single natural language request.
No YAML pipelines. No CI/CD config. Just tell your AI "deploy this" and it happens.
Related MCP server: Neptune MCP Server
Quick Start
# Run directly — no install needed
npx github:abbylester1/golive-mcp
# Or install globally
npm install -g github:abbylester1/golive-mcp
golive-mcpCreate a deploy.config.json in your project root, then tell your AI:
"Deploy to production"
Use Cases
You want to... | Go Live handles it |
Ship a fix fast | Commit, push, create PR, deploy to cPanel — one command |
Deploy to shared hosting | FTP upload to Spaceship CageFS + copy to real path + Passenger restart |
Deploy to Vercel | Build + |
Deploy to any VPS | Custom shell commands (rsync, scp, whatever you need) |
Verify it worked | Health check on your deploy target, returns status + body |
Let non-tech people deploy | They talk to the AI, the AI runs Go Live |
Multi-target pipeline | Deploy to staging (Vercel) and production (cPanel) from the same config |
Quick rollback | Switch branch, deploy again, verify the old version is gone |
Installation
# Run without installing
npx github:abbylester1/golive-mcp
# Or install globally
npm install -g github:abbylester1/golive-mcp
golive-mcpMCP Client Setup
claude mcp add golive-mcp -- npx github:abbylester1/golive-mcpAdd to .cursor/mcp.json:
{
"mcpServers": {
"golive-mcp": {
"command": "npx",
"args": ["github:abbylester1/golive-mcp"],
"env": {}
}
}
}Add to .vscode/mcp.json:
{
"servers": {
"golive-mcp": {
"type": "stdio",
"command": "npx",
"args": ["github:abbylester1/golive-mcp"]
}
}
}opencode mcp add golive-mcp -- npx github:abbylester1/golive-mcpWhat it does
Three tools your AI can call:
create_pr
Create a GitHub PR from your current branch.
{
"title": "Fix login bug",
"body": "Closes #42",
"base": "main",
"commitMessage": "fix: login redirect"
}If commitMessage is set, it auto-commits everything. Just need gh CLI installed.
deploy
Build and deploy to production (or any target you've configured).
{
"target": "production",
"skipBuild": false
}Works with any hosting provider.
status
Check if your site is live.
{
"target": "production"
}Hits your health check URL and reports back what it got.
Hosting you can deploy to
Provider | How Go Live deploys |
Spaceship (cPanel) | FTP over TLS to CageFS path, optional PHP copy to real path, Passenger restart |
Vercel | Runs |
Any cPanel / Plesk | Custom shell commands (rsync, curl, etc.) |
AWS EC2 |
|
AWS S3 + CloudFront |
|
AWS Elastic Beanstalk |
|
Google Cloud Run |
|
Google Compute Engine |
|
Azure App Service |
|
DigitalOcean App Platform |
|
DigitalOcean Droplet |
|
Fly.io |
|
Railway |
|
Netlify |
|
Render | API curl to trigger deploy |
Heroku |
|
Cloudflare Pages |
|
GitHub Pages | Push to |
Any VPS (Linode, Vultr, Hetzner, OVH) |
|
If Go Live doesn't have a built-in provider for your host, use the custom provider — it runs any shell commands you want.
Spaceship (cPanel shared hosting)
Go Live is the only MCP deploy tool that handles Spaceship's CageFS setup:
{
"targets": {
"production": {
"provider": "spaceship",
"build": { "command": "npm run build", "cwd": "frontend" },
"spaceship": {
"host": "server43.shared.spaceship.host",
"user": "user@domain.com",
"password": "",
"remotePath": "/app",
"cageFsPath": "/home/user/domain.com/kyro/app",
"phpCopyUrl": "https://domain.com/deploy.php"
},
"healthCheckUrl": "https://domain.com/api/health"
}
}
}Custom (any host not listed)
{
"targets": {
"production": {
"provider": "custom",
"build": { "command": "npm run build", "cwd": "." },
"custom": [
"rsync -avz --delete out/ user@host:/var/www/html/",
"ssh user@host 'systemctl reload nginx'"
],
"healthCheckUrl": "https://mysite.com/api/health"
}
}
}Configuration
Create deploy.config.json in your project root (or Go Live finds it here):
deploy.config.jsonin current directorydeploy.jsonin current directory~/.config/golive-mcp/config.json--config=/path/to/config.jsonflag
How it's different
vs | They do | Go Live does |
GitHub MCP servers | GitHub API operations only | GitHub + build + deploy + verify |
AppDeploy / NEXUS AI | Lock you into their cloud | Works with your hosting |
DeployPilot | Heavy gateway (91 tools, Docker) | Lightweight MCP server, single binary |
EZKeel / Runway | CLI tools, need a VPS | Works from any AI via MCP |
CI/CD pipelines | Need YAML config | One JSON config, AI manages the rest |
Why
Production deploys are the most error-prone part of AI-assisted coding. You write code, commit, push, create a PR, build, upload, copy files, restart, check if it's live — that's 9+ manual steps. Go Live turns it into one sentence: "deploy to production."
License
MIT
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/abbylester1/golive-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server