Getting Started

Introduction

Zeeb MCP is a hosted backend infrastructure platform that exposes a full suite of developer tools through the Model Context Protocol (MCP). It lets AI agents — such as Claude, Cursor, and any other MCP-compatible client — create and manage backend projects, generate code, run migrations, inspect databases, manage files, and more, all through a secure, multi-tenant API.

What you get

CapabilityDescription
Hosted projectsIsolated zeebpy project workspaces per tenant, with git and a live preview URL
50+ MCP toolsFirst-class tools for every stage of backend development
API key managementScoped, auditable API keys with revocation and per-key feature flags
Multi-tenant isolationEvery tenant's data, projects, and keys are strictly isolated
Usage meteringTool calls counted against plan limits; real-time usage tracking
OAuth & bearer authSupports both direct bearer token auth and interactive MCP OAuth

How it works

AI Agent (Claude / Cursor)
        │  MCP over HTTP
        ▼
  Zeeb Gateway  ──── Auth & quota check ──── Billing service
        │
        ▼
  Worker service  ──── Executes tool ──── Project workspace
        │
        ▼
  Result returned to agent
  1. Your AI agent connects to https://mcp.zeeb.cloud/mcp with an API key.
  2. The gateway authenticates the key, checks your plan quota, and creates a job.
  3. The worker executes the tool in your tenant's isolated workspace.
  4. The result is returned to the agent via the MCP protocol.

MCP server URL

https://mcp.zeeb.cloud/mcp

All MCP tools are mounted at this single endpoint. Authentication is via a Bearer token in the Authorization header.

Next steps