What is Conjra?

Conjra is an MCP (Model Context Protocol) server that gives AI-powered code editors — all 20 supported editors — the ability to provision real cloud infrastructure on your behalf.

Instead of leaving your editor to click through cloud dashboards, you simply ask your AI to set things up. Conjra handles the API calls, injects credentials, and wires everything together.

The problem it solves

Setting up cloud infrastructure traditionally requires:

• Navigating provider dashboards (Stripe, Supabase, Vercel, etc.)
• Generating and copying API keys
• Installing SDKs and configuring clients
• Managing environment variables
• Setting up webhooks and callbacks

Each provider has its own flow, its own portal, its own terminology. Conjra unifies all of them behind a single interface: natural language commands in your editor.

Who it is for

• Vibe coders who want to ship full-stack apps without context-switching
• AI-assisted developers using Claude Code, Cursor, Windsurf, or any of 20 supported editors
• Solo builders who hate clicking through dashboards
• Teams that want infrastructure setup codified and repeatable

Quick start

npm install -g conjra conjra init --ai claude conjra add supabase

Then in your editor, ask your AI: "use conjra to create a supabase project called my-app". Conjra provisions the resource, injects credentials, and returns the connection details.

Supported providers

Conjra works with 19 providers: Anthropic, AWS Amplify, Clerk, Cloudflare, Firebase, Fly.io, GitHub, Loops, Neon, OpenAI, PlanetScale, Railway, Replicate, Resend, Stripe, Supabase, Twilio, Upstash, and Vercel.

See the Providers section for details on each.