Inspiration

While building an AI to manage my social media, I needed a system prompt that matched my tone. Testing across DeepSeek, Claude, and ChatGPT was frustrating—nothing was user-friendly or smart. That’s why I built PromptingLab.

What it does

PromptingLab is a workspace to write, test, and improve system prompts using multiple LLMs and AI-assisted tools—all focused on clarity and iteration.

Proudest Feature

The Prompt Enhancer rewrites your system prompt based on output feedback. You just describe what’s wrong—no more manual fixing.

How we built it

Built with Bolt.new (UI), Supabase (backend), Netlify (hosting), and Entri (domain). Supports OpenAI, Anthropic, DeepSeek, and more via API.

Challenges

Supabase Edge functions were unreliable, which slowed us down. Still, we made it work without giving up.

Accomplishments

We solved a real pain point. I already use it daily, and we’re turning it into a full SaaS product post-hackathon.

What we learned

Prompt engineering is iterative. A fast, collaborative interface changes everything. Good UI and smart AI tools save hours.

What’s next

Selling it, adding team collaboration, support for more models, and custom LLM testing. This is just the start.

Bonus and sponsor challenges:

Creative use of AI PromptingLab transforms the way prompt engineers work enabling real-time, side-by-side testing across multiple LLM with AI-Assisted edits and performance metrics.

Uniquely Useful Tool Born from daily frustration, PromptingLab solves the real pain of writing and testing prompts. It’s purpose-built for prompt engineers—finally, a tool that makes iteration fast, trackable, and reusable.

Sharpest Problem Fit PromptingLab nails a clear need: prompt development is messy and slow. This tool unifies writing, testing, and refining into one clean loop—built by someone living the problem.

We Didn’t Know We Needed This PromptingLab quietly fixes what every prompter endures: switching tabs, losing context, repeating tests. It feels obvious once you use it—and hard to go back without it.

Future Unicorn PromptingLab is positioned at the heart of LLM tooling. It’s fast, focused, and developer-first—exactly the kind of vertical product that scales with the rise of AI workflows.

Inspirational Story As a full-time prompt engineer, I built PromptingLab after countless late nights of frustration. No budget, just need. I stayed up till sunrise to ship what I couldn’t find.

Most beautiful UI PromptingLab’s minimalist design is built for focus—clean layout, fluid motion, and light/dark modes that enhance flow without distractions.

Deploy challenge PromptingLab is fully deployed on Netlify with instant global delivery, zero-config hosting, and fast CI/CD. It’s a clean, production-ready frontend built the Netlify way.

Startup challenge Supabase powers all backend logic, user auth, secure API key storage, prompt history, and LLM calls via Edge Functions. We use Supabase exactly as intended, end-to-end.

Custom domain challenge PromptingLab runs on a custom domain with full HTTPS, DNS, and branding, professionally set up and integrated via Entri.

Built With

  • anthropic
  • bolt.new
  • deekseek
  • entri
  • gemini
  • netlify
  • node.js
  • openai
  • openrouter
  • stackblitz
  • supabase
Share this project:

Updates