This project is a prompt library that provides reusable prompt artifacts designed to override or test prior instructions of AI systems. It targets flagship AI models, though compatibility is not verified.
Project overview
The repository documents prompt artifacts specifically intended to instruct target AI systems to disregard previous instructions, targeting all flagship AI models.
Project type
Prompt Engineering
Deployment
Refer to project documentation
License
AGPL-3.0
Best for
Developers researching prompt-engineering techniques and instruction adherence in flagship AI models.
Key capabilities
Supplies reusable prompt artifacts described as liberation prompts for AI systems.
Contains prompt text that directs a target AI system to disregard previous instructions.
Documentation describes the prompt artifacts as targeting all flagship AI models without asserting verified compatibility.
Limitations and risks
The prompt content explicitly asks target AI systems to disregard previous instructions, creating adversarial prompt-use risk.
Runtime requirements including operating systems, GPU needs, minimum hardware, databases, external services, and model dependencies are not documented.
Getting started
The setup difficulty, prerequisites, and first success path for adopting these prompt artifacts are not documented.