artwhisper
Neo-Manga Cover

Neo-Manga Cover

Grok Imagine Concept ArtCharacter Design

Generation Prompt

[SUBJECT], [CONCEPT], [MOOD], [RED_ELEMENTS], [ASPECT]

Full neo-manga cover template — hard ink, angular shadows, saturated red on cream paper, dominant diagonal subject.

---

## Input Mode Handling
- **Text-only** (no reference image): Fill [VARS] inline, assemble via base template.
- **With reference image** (I2I via Grok): Use ref for character identity only — do NOT preserve background, pose, setting, or palette from reference. [VARS] control all style/composition aspects.
- **Skill-to-skill pipeline** (via imagent skill processor): Read neo-manga-cover SKILL.md, parse user's brief, auto-fill [VARS] from intent, regenerate polished output.

## Output Rules
- ONE prompt per generation — no alternatives, no commentary, no reasoning.
- No headings, no "Prompt:" label, no markdown wrapping in actual generation.
- Append exact suffix: "no signature, no watermark, no artist name, no text on image"

---

## Base Template (fill [TOKENS] below):

[SUBJECT], [CONCEPT], [MOOD], graphic neo manga illustration built from hard ink, angular shadows, saturated red accents, cream paper space, and abstract motion marks, feels like a printed manga cover plate mixed with fashion editorial illustration and graphic poster design, clean in shape, aggressive in line, and controlled in color, close dominant subject on strong diagonal, tilted compressed intentional frame, cropped edges, sweeping background strokes, angled posture, overlapping forms, sharp negative space, restricted palette of ink black deep charcoal blood red hot coral pale peach parchment cream muted gray small white highlights, precise manga ink lines sharp tapered varied contours, thin hairline strokes thick black shapes quick scratch marks for folds strands seams cracks motion, shadows as graphic cut shapes large black areas hard cel shaded transitions angular cast shadows thin highlight cuts jagged interior shapes, red as material and force in [RED_ELEMENTS], background as abstract field of motion cream paper crossed by coral strokes black diagonals halftone patches splattered ink scratched lines torn paint shapes speed marks broken graphic arcs, surfaces with flat tone ink texture screen tone dry scratches selective highlights, angular folds black pools in fabric, long pointed strands in hair smoke ribbons wires energy shapes, sharpest linework and color contrast at focal point eyes face hands edges of clothing key tools symbols, restrained halftone dotted fields fine grain imperfect ink coverage tiny specks rough paper texture manga screen tone only where supporting depth rhythm surface, [MOOD] through posture crop gaze direction diagonal pressure and contrast

## Surfaces & Materials
- Cream/parchment paper base — never white digital canvas
- Ink-on-paper texture: dry brush, splatter, scratch marks, imperfect coverage, tooth of paper
- Screen tone (halftone dots) only where supporting depth — restrained, fine-grain, imperfect
- Highlights: thin white/cream cuts, selective, never glossy or 3D
- Red behaves as material (ink pool, blood splatter, fabric dye, visor glow) — never as generic digital color fill

## Emotional Temperature via Contrast
- Deadly calm → high contrast, clean negative space, thin controlled lines
- Chaotic fury → heavy black areas, scratch lines, torn shapes, wide red pools
- Melancholic → muted reds, crease shadows, downward gaze, restrained palette
- Victorious → upward diagonal, white highlight cuts, thin crisp red accent lines

## Anti-Patterns (NEVER)
- Photorealism, glossy 3D, soft airbrushed shading, rounded cartoon softness, generic fantasy polish
- Neon cyberpunk excess, cluttered backgrounds, rainbow palettes, painterly mud, realistic lens effects
- Perfect symmetry, overdescribed anatomy, uniform outlines, random glitch, stock prompt phrases
- No signature, no watermark, no artist name, no text on image

---

## Variables
| Variable | Description | Example |
|----------|-------------|--------|
| [SUBJECT] | Main character, creature, object, or scene | femme fatale assassin / ancient warrior / fairy tale goddess |
| [CONCEPT] | Narrative hook, setting, or metaphor | cyberpunk city battle stand / porcelain throne of thorns |
| [MOOD] | Emotional temperature (see section above) | deadly calm / victorious / severe / chaotic fury |
| [RED_ELEMENTS] | Where red appears as material or force | red visor glow / flame weapon / bloodstained ribbons |
| [ASPECT] | Aspect ratio — 3:4 default | 3:4 / 9:16 / 1:1 / 16:9 |

---

## Self-Referential Notes
- This skill is designed for DeepSeek v4 Flash (brain) → Grok/Hailuo/Z-Image (hands) architecture.
- DeepSeek reads this SKILL.md, fills [VARS], assembles prompt, then calls the image gen MCP.
- The brain model NEVER generates the image — it only processes the skill and routes.
- For Hailuo fallback: same prompt, but Hailuo understands ink/paper textures less reliably — DeepSeek may add explicit "ink illustration on cream paper" reinforcement.

no signature, no watermark, no artist name, no text on image