Why we are featuring it
Looking inside the instructions that guide models.
CL4R1T4S is a useful collection for understanding how AI systems are steered. It compares prompts, rules, and documented behaviors, turning hard-to-find material into a resource people can inspect.
What is inside
- Prompts from OpenAI, Google, Anthropic, xAI, Perplexity, Cursor, and Devin.
- Documentation of what systems can say, refuse, or redirect.
- Persona and capability details imposed on models.
- Contributions organized with name, version, date, and context.