Happyhorse Prompter

The core discipline is subject-first grammar, explicit camera specs, and audio activation phrases like "with rain on leaves audible" or "speaking English at natural pace". It includes an 8-second temporal beat structure for sequenced shots. The hard rule on duration placement is useful: either put it at the very front as a timing anchor or omit it entirely from the prompt and note it in metadata. If you need multi-shot sequences or 10-15 second complexity, the docs point you to the Seedance skills instead. Good for coffee shop ambience, ASMR prompts, or any 3-8 second clip where you want the model to generate matching sound without post work.