Universal Prompt Designer

It started with 147 prompts. I was recreating a prototype in Figma Make, checked my log expecting a few dozen, and found 147 before I was a third through the flow. The tool wasn't burning tokens. My prompting was. So I looked at the prompting tools already out there. Most fell into two patterns. Some were prompt expanders that added structure by inferring my intent for me. Others pushed the clarification work back onto me, asking me to decide before they'd generate anything. Several needed a subscription, a download, or a browser extension just to get started. Figma Make was already inferring plenty on my behalf. I didn't want more inference stacked on top, and I didn't want to be handed the clarification and left to sort it out alone. So I built the tool I actually wanted. Universal Prompt Designer doesn't expand your idea and it doesn't hand the work back to you. It guides you through clarifying the idea, asks targeted questions about what you're building, and turns your answers into a complete, structured prompt. No template to fill in. No prompt engineering knowledge required. The output follows an 8-part framework: role and context, purpose and goal, inputs, instructions and logic steps, output requirements, constraints, fallback logic, and tone guidance. That structure gives the AI enough to stay on brief across a full session, not just a single response, so the prompt is ready to run the first time instead of a draft you keep correcting. Most sessions take 3 to 7 credits. Bring more context upfront and the session gets shorter. The finished prompt pastes directly into ChatGPT, Claude, Gemini, Cursor, v0, Perplexity, NotebookLM, Figma Make, or whatever the project calls for. A companion tool, Visual Prompt Architect, handles image and video prompts separately. Same interview format, different vocabulary, built around how image models process input.