Professional art drawing and illustration analysis. Use when the user asks to analyze an artwork, reference image, character art, manga/anime/manhwa panel, concept art, sketch, painting, or generated image with focus on linework, drawing technique, face proportion fingerprint, facial-feature construction, hair construction, strand density, medium/material diagnosis, traditional hand-drawn vs digital art cues, rendering workflow, color handling, glow/specular profile, accessory/detail rendering, imperfection/polish profile, material feel, brushwork, visual style, art school/tradition, composition, style-transfer fidelity, and how to recreate or prompt the style in technical detail.
Generate mobile portrait character images from a reference image plus a character description. Use when the user wants imagegen to strictly follow the reference art style, face proportion fingerprint, facial-feature construction, linework fingerprint, visible medium/material behavior, hand-drawn or digital art cues, pose, facial expression, lighting, glow/specular behavior, rendering, eye rendering/coloring, hair construction/coloring, restrained strand density, imperfection/polish level, detail density, palette handling, accessory/detail rendering, and background style while replacing the reference identity with the supplied character/OC appearance. Applies analyze-art-drawing style analysis, trait translation, 9:16 vertical framing, safe focal placement, visual-budget control, makeup blending, background-content separation, prompt construction, QC, and one targeted reroll.
Recast or patch a supplied character reference image using a character prompt while preserving source pose, crop, face construction, expression, hand/crop logic, apparent scale, linework, lighting, rendering model, medium behavior, style-family balance, detail density, polish ceiling, hair construction, grain/noise behavior, and background style. Use when imagegen should edit a reference character by changing only local identity traits, colors, visible marks, outfit cues, accessories, and compressed background content instead of redrawing from scratch. Supports source-crop and Mobile portrait 9:16 modes with app-safe focal placement, canvas extension without poster redraw, targeted conditional guards for eye/hand/hair/background risks, strict art-fidelity QC, and hard basic-anatomy failure gates. Triggers include make this into my OC, change outfit, patch traits, keep art/pose/face but replace hair/eyes/clothes/accessories/background, mobile portrait, app avatar, 9:16, and avoid fantasy-glowy redraw drift.
Compose a new original character from a pasted reference character card, prose sheet, or messy mixed text. Use when the user asks to create, remix, derive, or compose a new character from an existing character while returning a database-ready Mero character JSON object plus an IMAGE_RECAST_HANDOFF for recast-character-in-reference image prompting. Supports optional idea/hint text such as eye colors, outfit direction, hair, species, vibe, or concept constraints; preserves source gender while changing appearance, outfit, species traits, and copyrighted identity enough to avoid near-copying. The image handoff is always Mobile portrait mode / 9:16 and must not invent face scars, face marks, facial tattoos, birthmarks, or under-eye symbols unless the user explicitly asks for them.