Defines image types, visual rules, and prompt patterns for brand-consistent image generation
Define exactly what each type of image should look like before any prompt is written. This skill is the visual equivalent of a content brief — it ensures every image is intentional.
{concept description}. Clean, modern design with {brand_colors}. Single focal composition. Professional quality. {platform} post format.Bold cover image for carousel about {topic}. {brand_colors} background with space for large headline text. Premium, modern design. Eye-catching and clean.Clean mockup of {product_name} interface on {device}. Showing {feature_description}. Modern UI design with {brand_colors}. Realistic screen content.Modern, abstract visualization of {concept}. Premium, clean design. {brand_colors} color scheme. Intelligent and sophisticated feel. No text unless specified.Clean product explainer visual showing {feature/workflow}. Step-by-step flow or feature grid. {brand_colors}. Professional and scannable.Clean, premium card design for founder quote. {brand_colors} accent. Space for text overlay. Professional, personal, trustworthy feel. Minimal design.strategy/brand-knowledge-base.jsonstrategy/product-truth/{product}.mdThis skill's rules are codified in strategy/brand-knowledge-base.json under image_types.
The enrich_image_prompt.py script reads these rules automatically when --image-type is specified.
For manual use: reference this skill when writing image briefs to ensure correct visual direction.
HONEST CLASSIFICATION: This is a prompt template, not an executable plugin. OpenClaw injects this as context to guide LLM behavior. It does NOT enforce rules at runtime.