🎯 PromptCraft: AI Prompt Engineering Assistant

PromptCraft is an intelligent, web-based assistant designed to help you transform your ideas into powerful, effective, and well-structured prompts for generative AI models. It bridges the gap between a simple goal and a high-quality prompt that yields superior results.

The application leverages a Retrieval-Augmented Generation (RAG) system, which learns from user-rated prompts stored in a local knowledge base. Over time, as more prompts are generated and rated, PromptCraft becomes smarter and provides even better suggestions tailored to similar goals.

✨ Key Features

• Intelligent Prompt Generation: Describe your goal in plain English, and PromptCraft will generate three distinct, high-quality prompt variations.
• Flexible AI Model Selection: Choose from a curated list of the latest stable Gemini models or enter a custom model ID to experiment with preview or unlisted versions.
• Retrieval-Augmented Generation (RAG): Uses a knowledge base of previously successful, high-rated prompts to inform the generation of new ones.
• User Intent Analysis: Automatically analyzes your goal to determine its category, complexity, and key tags.
• Feedback Loop: Rate the generated prompts to help the system learn. Your feedback directly improves future suggestions for everyone.
• Local Knowledge Base: All prompts and ratings are stored locally in a SQLite database (promptcraft.db), giving you full control over your data.

🚀 Getting Started

Follow these instructions to set up and run PromptCraft on your local machine.

Prerequisites

• Python 3.10
• A Google Gemini API Key. You can get a free key from Google AI Studio.

Installation

1. Clone the repository:

   git clone https://github.com/your-username/promptcraft.git
   cd promptcraft

2. Create and activate a virtual environment (recommended):

   # For Unix/macOS
   python3 -m venv venv
   source venv/bin/activate
   
   # For Windows
   python -m venv venv
   .\venv\Scripts\activate

3. Install the required dependencies:

   pip install -r requirements.txt

4. Run the Streamlit application:

   streamlit run app.py

5. Open your web browser and navigate to the local URL provided by Streamlit (usually http://localhost:8501).

💻 Usage

1. Enter API Key: Paste your Google Gemini API Key into the input field in the sidebar.
2. Select a Model: Choose your desired Gemini model from the dropdown.
   • Standard Models: Select from the list for a balance of power and stability (e.g., gemini-1.5-pro-latest, gemini-1.5-flash-latest).
   • Custom Models: To use a preview or unlisted model, select "Enter Custom Model..." and type the exact model ID into the text field that appears.
3. Define Your Goal: In the main text area, describe what you want the AI to do.
4. Generate Prompts: Click the 🚀 Generate Prompts button.
5. Review Variations: PromptCraft will produce three distinct prompt variations.
6. Copy and Test: Use the 📋 Copy button to copy a prompt.
7. Rate the Prompts: Click the ⭐ Rate button and provide feedback to improve the system.

🤔 FAQ

Q: Why would I use a custom model name? A: Google often releases new models in a "preview" state before they are generally available. These models might have new features or improved performance. The custom model input allows you to experiment with these new models (e.g., gemini-2.5-pro-preview-08-01) as soon as they are announced, without waiting for an application update.

Q: What's the difference between the listed Gemini models? A: The models offer a trade-off between performance, speed, and cost. gemini-2.0-pro is highly capable for complex tasks, while gemini-2.0-flash is optimized for speed and cost-efficiency.

Q: Where is my data (prompts, ratings) stored? A: All data is stored locally in a file named promptcraft.db in the project's root directory. You have full ownership and control over this data.

🤝 Contributing

Contributions are what make the open-source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is distributed under the MIT License.