what is prompt-patterns fr?

Prompt Patterns is an open guide that teaches you how to write better instructions for AI tools like ChatGPT and Stable Diffusion. Instead of randomly typing requests and hoping for good results, it introduces reusable "patterns", borrowed from the world of software design patterns, that help you structure your prompts so the AI reliably does what you want. The project also links to a companion tool called ClickPrompt that integrates these patterns. The guide walks through several core patterns. "Specific instruction" means asking a direct question or giving a keyword. "Instruction template" gives the AI a clear format to follow, like summarizing text using the STAR method. "Proxy" asks the AI to act as someone else, a customer service agent, a Linux terminal, whatever role fits. "Demonstration" shows the AI examples of what you want so it can mimic the pattern. Beyond these basics, the guide covers techniques like using symbolic rules, negative prompts (telling the AI what to avoid), and iterative refinement where you use the AI's own output to improve the next prompt. This resource is for anyone working with AI text or image generation who wants more consistent, controllable results. A product manager could use the template pattern to get structured competitive analyses. A developer could use proxy mode to make ChatGPT behave like a code console. A marketer could use demonstration mode to generate product descriptions matching a house style. The patterns are framework-agnostic and work across tools. What makes the project notable is its conceptual approach: it treats prompt writing not as ad hoc magic but as a discipline with repeatable structures you can learn, combine, and teach. Some sections are still works in progress, and the guide is openly soliciting contributions, translations, and improvements from the community.