Prompt Generator

This project aims to accelerate the process of generating prompts for Large Language Models (LLMs). Users can input their intention, choose a development mode, and the software will automatically generate high-quality prompts. It also includes features for clarifying requirements and connecting to various LLM APIs.

Features (Planned)

• User Intent Input: Capture user's project intention.
• Development Modes: Support iterative, one-shot, and other custom development modes.
• Requirement Clarification: Ask high-quality questions to refine user's needs.
• High-Performance Security Engine: Advanced scanning for PII, prompt injection, and other threats with strict error handling and sanitization.
• LLM API Integration: Connect to Gemini, Claude, OpenAI, Deepseek, etc.
• Prompt Template Library: Save and reuse custom prompt templates.
• Interactive Prompt Refinement: Allow users to interactively improve generated prompts.

Setup

1. Clone the repository:

   git clone <repository_url>
   cd prompt_generator

2. Create and activate a Conda environment:

   conda create -n promptgen python=3.9
   conda activate promptgen

3. Install dependencies:

   pip install -r requirements.txt

Usage (Planned)

python src/main.py generate "create a simple web server with Python"

Development Report

The development report will be located in the docs/ directory.