h4cker/ai_research/prompt_engineering/bug_bounty_prompt_generator
2024-09-10 23:08:51 -04:00
..
prompts Add files via upload 2024-09-10 23:08:51 -04:00
ai_prompt_maker.py Add files via upload 2024-09-10 23:08:51 -04:00
README.md Create README.md 2024-09-10 23:07:59 -04:00
top_vulns.txt Add files via upload 2024-09-10 23:08:51 -04:00

AI Prompt Generator for Bug Bounty Vulnerabilities

This script generates AI-powered prompts for different vulnerabilities using OpenAI's GPT model. It's designed to assist ethical hackers and bug bounty hunters in creating detailed and specific prompts for different types of vulnerabilities.

Features

  • Reads a list of vulnerabilities from a file
  • Generates AI-powered prompts for each vulnerability
  • Saves each generated prompt to a separate file
  • Uses OpenAI's GPT model via LangChain
  • Implements a custom processing chain for prompt generation

Prerequisites

  • Python 3.6+
  • OpenAI API key (set in .env file)
  • Required Python packages (see requirements.txt)

Setup

  1. Clone this repository
  2. Install required packages:
    pip install -r requirements.txt
    
  3. Create a .env file in the project root and add your OpenAI API key:
    OPENAI_API_KEY=your_api_key_here
    
  4. Create a top_vulns.txt file in the project root, listing one vulnerability per line

Usage

  1. Ensure your top_vulns.txt file is populated with the vulnerabilities you want to generate prompts for
  2. Run the script:
    python ai_prompt_maker.py
    
  3. The script will process each vulnerability and generate a prompt
  4. Generated prompts will be saved in the prompts/ directory, with filenames based on the vulnerability names

Output

  • Console output will show each vulnerability being processed, the generated prompt, and the file where it was saved
  • Generated prompts will be saved as text files in the prompts/ directory
  • Each prompt file will include a word count and the uppercase version of the prompt

Customization

You can modify the system prompt in the script to adjust the style or focus of the generated prompts. Look for the prompt_template variable in the script.