One of the biggest projects developed by our team is a groupware platform. What makes it unique is that we wrote only the frontend. Every part of the backend, from data storage to authentication, is handled by our own system: Skapi.
In this article, we’ll walk you through how we implement security directly from the frontend using Skapi for backend. This isn’t theory, we’re sharing what we’ve learned from real projects we did ourselves.
What started as an experiment in prompt engineering turned into a working web app.
I wanted to see how far I could go using ChatGPT + Skapi (a zero-setup backend) to build a functional, Instagram-style application — no backend code, no server setup, no frameworks. Just prompts.
Turns out: you can do it in 2 prompts, if you give the AI the right documentation.
Most LLMs only read the first page of a URL, so we created a minimal, AI-optimized doc page that teaches ChatGPT how to use Skapi’s backend features: auth, database, real-time logic, etc.
The result: ChatGPT generated a working frontend + backend logic, with no manual intervention.
Curious if anyone else has tried similar workflows or tools.
