LogoAISecKit
icon of deck-of-many-prompts

deck-of-many-prompts

Manual Prompt Injection / Red Teaming Tool for large language models.

Introduction

Deck of Many Prompts

The Deck of Many Prompts serves as a manual prompt injection tool designed for Red Teaming purposes. This tool allows users to explore, manipulate, and gain insights into the behavior of large language models (LLMs).

Key Features:
  • Manual Prompt Injection: Effortlessly test and manipulate LLM outputs based on different prompts.
  • Tokenizer Demos: Demonstrates how various tokenization methods work, enabling deeper understanding of LLM mechanics.
  • Multiple Transformations: Includes a variety of transformations such as morse, braille, ASCII, base64, and more.
  • Installation Script: Simple installation process to get up and running quickly with clear commands to clone and set up the environment.
Benefits:
  • Enhanced testing capabilities for security professionals.
  • Aids in understanding the complexities and functionalities of AI and LLMs.
  • Encourages contributions from the community for tools engaged in ethical research and development.
Highlights:
  • Visit the live demo to see the tool in action!
  • Participate in the community by providing feedback and contributing to the ongoing development.

Newsletter

Join the Community

Subscribe to our newsletter for the latest news and updates