Pet Movie Poster Generator

Problem

A pet movie poster is harder than a simple costume swap: a standalone hero image may lack title typography, cinematic lighting, and readable poster hierarchy. Letting a model generate everything in one pass can also produce text overlapping the pet, identity drift, or a result that feels like a meme or social card instead of a film poster. This skill treats the final poster, not just a character image, as the target output.

How It Works

• Two-stage pipeline: call the image-process API first to turn a pet photo into a themed hero image, then call the design-generate API to compose the final poster.
• Identity preservation: stage 1 keeps species, fur pattern, face shape, ear shape, and body type while changing costume, props, environment, and lighting.
• Layout constraints: stage 2 treats the pet silhouette as a no-text zone, keeping the title, subtitle, and tagline in a dedicated safe area away from ears, whiskers, shoulders, and body edges.
• Theme presets: supports superhero, astronaut, medieval-noble, and closely related roles such as detective or knight.
• Quality checks: if the pet no longer looks recognizable, the poster lacks cinematic feel, or text still overlaps the subject, the workflow retries or adjusts the composition.

Boundaries

A public image URL is the preferred input; local attachments are only a temporary fallback and introduce privacy tradeoffs. The skill is best for a single cinematic pet poster, not bulk generation, video thumbnails, memes, e-commerce main images, or brand-exact typography work.