Anime Social Portrait Generator

Problem

Turning a real photo into a social avatar, Moments cover, or Twitter/X header often fails in two ways: the anime style is weak, or the face, hair, outfit, and body proportions become distorted. This skill targets photo-to-anime social visuals, aiming to preserve recognizable user traits while producing a more complete stylized social asset.

How It Works

• Input constraints: A valid mew.design API key and at least one clear photo are required; if either is missing, the flow pauses to ask for the missing input.
• Style inference: It selects the closest style from face shape, hairstyle, outfit, palette, pose, and overall vibe, such as energetic anime, Ghibli-inspired fresh style, Makoto Shinkai-like visuals, cyberpunk, guofeng, or cute chibi styles.
• Generation rules: It uses image/process, keeps avatar aspect ratio flexible by default, uses 16:9 for Moments covers, and 21:9 for Twitter/X headers.
• Quality checks: After generation, it checks style match, identity consistency, and natural proportions; weak results may be retried up to two times with stronger style keywords, scene cues, and identity-preservation constraints.

Scope

It works best when a usable personal photo is available and the goal is an avatar or social cover. It may not cover blurry photos, heavy facial occlusion, batch generation, text layout editing, or complex poster design. The output favors the person subject and stylized visuals, with no guarantee of exact brand reproduction, precise typography, or elaborate scene composition.