asgeirtj/system_prompts_leaks

This project serves as a valuable open-source data repository, meticulously compiling system prompts, messages, and developer instructions extracted from a wide array of popular AI models like ChatGPT, Claude, Gemini, Grok, and Perplexity. It offers a unique insight into the underlying directives that govern these advanced AI systems, providing researchers, developers, and security professionals with critical data for prompt engineering, understanding AI behavior, and identifying potential vulnerabilities or biases. The collection is regularly updated, ensuring relevance and comprehensiveness for the rapidly evolving AI landscape.