Yes. The community's most-shared design asset is a system prompt that tells the model to act like a senior graphic designer and to output one clean, copy-paste image prompt, instead of drifting into AI defaults. The idea is simple: human design is restraint, hierarchy, and deliberate choices, so the prompt forces the model to make choices.
It works by refusing the usual AI tells. Before it writes anything, it decides on one visual idea, one dominant element, what stays quiet, and one design tradition to commit to. Then it designs around your actual text rather than leaving generic poster space.
You do not have to memorize the theory. Paste a short anti-AI block like this at the end of any flyer prompt:
Asymmetric composition unless solemn symmetry required. One dominant read, one secondary read, some elements deliberately quiet. Design around the actual text, no fake poster space. Imperfect directional lighting, material-specific textures, one deliberate flaw. Commit to one specific design tradition. No synthetic parchment, no muddy grain wash, no faux vintage distress, no glow, no HDR, no over-sharpening, no bevels, no embossing, no canned flourishes. Empty space must have tension and relationship to the rest of the composition.
Then run the vision-check loop: generate the flyer, upload it back to the model, ask what is making this look like AI, and paste the block again with a request to improve based on it. Two or three passes usually get there.
The full system prompt, its style library, and worked examples are on the community wiki: The Designer Prompt.
Discussions are coming soon. Every skill, app, and guide will carry its own thread where shluchim compare notes on what worked.