START

Getting started

Yeeted hosts your AI-built app: database, logs, domains, the lot. You never touch a server: just tell your agent to deploy.

There are two ways to get going:

1. Connect your agent

Point Claude (or any MCP agent) at Yeeted and say “deploy my app.” Paste this into your agent's MCP config, then log in when prompted to authorize the connection:

2. Open the dashboard

During the beta, sign-up requires an invite code and email verification. The dashboard lets you manage deploys, databases, logs, and credentials.

Create an account

// your agent's MCP config
{
  "mcpServers": {
    "yeeted": { "type": "http", "url": "https://mcp.yeeted.ai/mcp" }
  }
}

Your agent registers its connection automatically and opens a browser where you log in to authorize it. That connection does not create a Yeeted account: during the beta, account creation needs an invite code and verified email. For CI or scripts, create an OAuth credential in the dashboard instead.

The login starts as soon as the client connects. Claude Code: run /mcp, pick yeeted and choose Authenticate. Claude Desktop and claude.ai: Settings > Connectors > Add custom connector, then Connect. Cursor: press Login next to yeeted in its MCP settings. VS Code: allow the “wants to authenticate” prompt. Before logging in, the tool catalog is public at https://mcp.yeeted.ai/.well-known/mcp/server-card.json. An agent with no account registers over HTTP (POST https://mcp.yeeted.ai/oauth/agent/request-code, then POST https://mcp.yeeted.ai/oauth/agent/register; see auth.md).

Your first deploy

Once your agent is connected, ask it to deploy. It sends your code to Yeeted, which builds it and hands back a live preview URL. When the preview looks right, ask it to ship to production. Deploying explains what happens on each step.

// just tell your agent
"Deploy my app to Yeeted."
"Ship this to production."
"Roll back to the last working version."