wrenshipDeploy from anywhere. Automate everything.
MCP tools for your AI assistant, a one-command CLI, Git auto-deploys and a GitHub Action.
Deploy channels
Six ways in.
Six ways in.
One pipeline.
MCPAI appsClaude, ChatGPT, Cursor, Claude Code, Codex and any MCP client.
WEBDashboard uploadDrop a folder or a ZIP.
GITGitHub & GitLabAuto deploy on push. Pull request previews on Edge Pro.
CLIwrenship deployOne command from any project folder.
CIGitHub ActionsDeploy from any workflow with one step.
AGENTAI coding agentsAgents with a terminal run
npx wrenship deploy --json.
MCP server
Let your AI run the deploy.
One remote MCP server over Streamable HTTP at https://api.wrenship.com/mcp. Sign in once with OAuth, then use it from any AI app that supports connectors. Setup for each app →
| Tool | What it does |
|---|---|
list_projects · get_project | Your projects with their live URL, latest build and domains |
deploy | Build and publish files, or redeploy a Git project. A preview deploy gets a temporary URL |
get_deployment · get_logs | Follow a build until it's live, and read its log so the AI can debug for you |
rollback | Put an earlier build live again |
list_env · set_env | List and set environment variables. Secret values are never returned |
add_domain · check_domain | Connect a domain and check its DNS and SSL |
get_analytics | Visitors, top pages, response times and errors |
Quick start
Copy, paste, deployed.
CLI
$ npx wrenship deploy ◇ Project ./toko-batik (Vite + React) ◇ Build done in 21s ◇ Upload 38 files ✓ Live https://toko-batik.wrenship.site
MCP connector
{
"mcpServers": {
"wrenship": {
"type": "http",
"url": "https://api.wrenship.com/mcp"
}
}
}CI with a token
$ WRENSHIP_TOKEN=wrs_pat_… npx wrenship deploy --json { "ok": true, "status": "ready", "url": "https://toko-batik.wrenship.site" }
GitHub Actions
- uses: wrenship/deploy-action@v1 with: token: ${{ secrets.WRENSHIP_TOKEN }}
Deploy from your tools.
Connect an AI app or run the CLI, and your first site is live in minutes.