What problem does it solve? Generating a cast of character portraits one at a time produces inconsistent results — different lighting, angles, and rendering styles that make characters look like they come from different games. This Skill enforces a locked visual anchor so every portrait in a cast shares the same framing, lighting, background, and style. ## Core Features & Use Cases - Locked Anchor System: Defines a reusable framing/lighting/style anchor in .summer/portrait-anchor.md that is appended verbatim to every portrait prompt for cast-wide consistency. - Identity-Preserving Iteration: Uses img2img with referenceImageUrl to tweak expressions or costumes without identity drift, instead of regenerating from scratch. - Engine Integration: Imports generated images via summer_import_from_url and wires them into dialogue UI TextureRect nodes with summer_set_prop. - Use Case: You need dialogue portraits for four party members in a visual-novel-style RPG. The Skill locks the anchor on the first portrait, then generates the remaining three with identical composition, and wires each into your dialogue scene. ## Quick Start Ask your agent to generate a VN-style dialogue portrait for a named character and lock the portrait anchor for the rest of the cast.
