ink-wash-poster

Invoke $ink-wash-poster followed by a natural-language description of your theme, mood, or subject. The skill develops the art direction, writes an image prompt, generates the bitmap, and returns the image with the final prompt and a short design rationale.

Attach the photo and tell the skill which features must stay recognizable, such as a ridgeline, pose, or silhouette. It preserves those key structures at a readable scale while dissolving peripheral edges into ink and paper rather than applying a uniform filter.

Yes. Provide the exact wording and placement, and the skill renders only that approved text. It never invents dates, years, titles, translations, or decorative characters, and it inventories visible text after generation to catch errors.

Yes. Add a phrase like "return the prompt only" to your request. The skill then stops after writing the four-part image-generation prompt instead of producing the bitmap, which is its default behavior.

Traditional Ink Mode activates only when you explicitly request traditional Chinese painting, guofeng, shanshui, calligraphy, or classical composition. Otherwise the default is contemporary minimal editorial design, with traditional motifs appearing only as small supporting fragments.

Image models can distort lettering, especially long copy. The skill performs one focused regeneration on text failures, but for long passages, brand-critical wording, or legally sensitive text, final typography should be added in a layout tool after generation.