PromptGen - AI Prompt Generator & Optimizer

This project is a full-stack application that allows users to generate, optimize, and manage AI prompts using popular AI providers (e.g., Gemini, GPT, Claude). It provides a modern frontend built with React + Material UI and a backend powered by FastAPI. Users can generate optimized prompts, view their history, and copy results easily.

Features

• Generate Prompts – Create prompts for different purposes like Blog, Social Media, Code, etc.
• Optimize Prompts – Improve existing prompts with AI and get multiple variants with scores.
• History Management – Save prompt history in browser localStorage.
• Modern Dark UI – Built with React and Material UI for a sleek user experience.
• One-Click Copy – Copy any generated/optimized prompt with a single click.
• Fast Backend – Built using FastAPI with endpoints for generation & optimization.

Tech Stacks

Frontend: React, Material UI, LocalStorage(For History)

Backend: Python, FastAPI, Unicorn

Large Language Model: Gemini/GPT/Claude

Project Structure

📦 prompt-app
 ┣ 📂 prompt-frontend   # React app (Material UI)
 ┃ ┣ 📂 src
 ┃ ┃ ┣ 📂 components  # UI components (Generate, Optimize, History)
 ┃ ┃ ┣ App.jsx        # Main App
 ┃ ┃ ┗ index.jsx
 ┃ ┗ package.json
 ┣ 📂 prompt-backend    # FastAPI app
 ┃ ┣ main.py     # FastAPI routes
 ┃ ┣ requirements.txt
 ┣ README.md

Setup & Installation

Backend Setup

cd prompt-backend
python -m venv venv
source venv/bin/activate   # On Windows: venv\Scripts\activate
pip install -r requirements.txt
uvicorn main:app --reload

Frontend Setup

cd prompt-frontend
npm install
npm run dev / start

API Reference

Generate Prompt

  POST /generate

{
  "topic": "I want to make aws cloud price optimizer using ai so gives the free resources and step by step process implementation.",
  "mode": "text",
  "use_for": "Normal",
  "tone": "Professional",
  "platform": "ChatGPT",
  "format": "Plain Text",
  "length": "Large",
  "strictness": "Loose"
}

Optimize Prompt

POST /optimize

{
  "user_prompt": "Post for LinkedIn jha me web dev ke liye roadmap de sku",
  "num_variants": 2
}

Future Improvements

• Add authentication & user accounts

• Store history in database instead of localStorage

• Deploy to cloud (AWS/GCP/Render/Vercel)

• Advanced analytics (prompt effectiveness scoring)

• Add Image and Video prompt generation

Authors

• @ashishalf

Contributing

Contributions are always welcome!

• Fork the repo

• Create a new branch (feature/your-feature)

• Commit changes

• Push & create a PR

License

MIT