visual-prompt-builder

You can generate IP-free visual prompts by processing narrative beats through this Skill, which strips copyrighted names and replaces elements with physical descriptions for photoreal text-to-image generation.

To ensure persistent visual state across storyboard beat batches, this Skill preserves physical descriptions and character states throughout single or batch processing. It maintains consistency by tracking characters_seen and enforcing strict JSON output across all narrative beats.

Yes, the visual prompt output works with FLUX and SeeDream pipelines. The Skill produces a strict JSON object containing the structured prompt and characters_seen list, specifically designed to integrate into these text-to-image generation pipelines.

You can generate photoreal prompts without copyright issues by applying IP-masking to your narrative beats. The Skill removes all IP names and substitutes them with precise physical descriptions, ensuring the visual output remains completely IP-free.

Yes, there are word count limitations for climax prompts in text-to-image generation. Standard prompts are limited to 60 words, while climax prompts allow up to 90 words, and every prompt must end with the designated storyRegister.