The EmPrompt App

⚡ Speed That Feels Like Cheating

Here's how fast it works:

1. Pick the AI Model to use
2. Pick a domain from the picker (1 click)
3. Type or paste your rough idea (as little as 5 characters, we dare you to try it — though we recommend more characters/words to get more precise results)
4. Hit Generate

EmPrompt fires your idea to the AI model of your choice, 2 minutes later — boom — a fully formed JSON prompt lands in the output panel. Copy it, download it, feed it directly into your pipeline.

What used to take 15 minutes of manual formatting now takes about 4/5 minutes.

📋 What You Get

- ✅ 18+ specialised prompt domains
- ✅ Structured JSON output (copy, download, or view as human-readable "TOON" format)
- ✅ Multiple AI model support (cloud)
- ✅ File upload support (.txt, .json) for batch ideas
- ✅ Token usage tracking per session
- ✅ Zero-data-collection privacy model
- ✅ Open-source ethos, extensible architecture

No more staring at a blinking cursor. No more "uhh, what lighting should I use?" No more copy-pasting the same scaffolding for the hundredth time.

The Technical Stuff

What is JSON prompting?

JSON prompting is the practice of expressing an AI prompt as structured JSON (such as key/value pairs and nested objects) rather than plain text, so a downstream system or model can parse the fields, enforce constraints, and produce structured outputs.

It separates intent, context, instructions, and desired output schema to make prompts machine-readable and easier to validate, transform, or reuse.