GitHub radar
Gathered Scenes Zine: Photo to Art with GPT Image 2
Two Codex skills that turn an ordinary photo into a handcrafted zine-style poster — either a real-photo collage or a fully new illustration distilled from the scene's mood and tension.
Gathered Scenes Zine is a collection of two Codex skills for AI-powered photo-to-art generation using GPT Image 2. The first skill (Gathered Scenes, `$scenes-gathered-zine-v1-3`) keeps the original photo as a visual anchor in a collage that adds abstract shapes derived from the photo, a single high-saturation colour, and visible torn-paper fibre edges. The second skill (Scene Distillation, `$scene-distillation-zine-v1-3`) does not include the photo in the result — it extracts the semantic core, emotional tension, and visual metaphor, then creates an entirely new illustration in an editorial minimal style. Skills are installed to `~/.codex/skills/` and invoked inline: `$scenes-gathered-zine-v1-3 [description]`. Repository includes a before/after case archive showing exactly what was preserved from each original scene. Released roughly six weeks ago and reached 4,000+ stars. Requires OpenAI Codex with GPT Image 2 access.
Why a vibe-coder should care
If you use OpenAI Codex and want to turn photos into art-quality posters without manually engineering image prompts, these skills handle the visual decision-making — what to keep, what to distil, how to balance the layout — through a structured Codex workflow.
How to install
Copy this and send it to your agent — Claude Code, Codex, any of them:
Install the zine poster skill from https://github.com/Zeejay0/gathered-scenes-zine-skill — clone the repo, copy skills/scenes-gathered-zine-v1-3 to ~/.codex/skills/, restart Codex, and show me how to use $scenes-gathered-zine-v1-3 on a test photo.
▌ More finds