liminal-editorial-posters

KO | EN

gitlite — search

#machine-learning #openai #awesome #gemini #devops #kotlin #privacy #chatgpt #large-language-models #terminal

liminal-editorial-posters

(★ 9)

A Codex skill for premium 3:4 split editorial photo posters with adaptive Liminal Frame compositions.

파일 탐색기

ZIP 다운로드

- openai.yaml
- hong-kong-victoria-harbour.png
- london-westminster.png
- san-francisco-golden-gate.png
- shanghai-pudong.png
.gitignore
README.md
SKILL.md

# CDN으로 사용하기

jsDelivr

jsDelivr는 공개 GitHub 리포지토리를 별도 설정 없이 CDN으로 즉시 서빙합니다. 버전과 파일을 고르면 웹페이지에 바로 붙일 수 있는 링크와 예시 코드가 만들어집니다.

버전

파일

링크

예시

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

A Codex skill for turning each uploaded photograph into an independent premium 3:4 editorial poster.

The top half preserves the source photograph. The bottom half reconstructs its defining subject or connected narrative ensemble through a restrained geometric threshold, frame-crossing photorealism, intentional negative space, and refined typography.

## Highlights

- One independent poster per source photo
- Exact 3:4 portrait canvas with a strict 50/50 split
- Faithful photographic preservation in the top panel
- Relationship-aware subject reconstruction in the bottom panel
- Adaptive horizontal, vertical, or diagonal geometric window
- Source-derived color, supporting marks, and editorial copy
- Output inspection with one targeted correction pass for hard failures

## Examples

Each source photograph becomes its own poster. These examples show how the same visual system adapts to different skylines, landmarks, colors, and spatial rhythms.

<table>
 <tr>
 <td align="center" width="50%">
 <img src="examples/san-francisco-golden-gate.png" alt="San Francisco Golden Gate Bridge editorial poster" width="100%"> 
 San Francisco · Golden Gate Bridge
 </td>
 <td align="center" width="50%">
 <img src="examples/shanghai-pudong.png" alt="Shanghai Pudong skyline editorial poster" width="100%"> 
 Shanghai · Pudong
 </td>
 </tr>
 <tr>
 <td align="center" width="50%">
 <img src="examples/hong-kong-victoria-harbour.png" alt="Hong Kong Victoria Harbour editorial poster" width="100%"> 
 Hong Kong · Victoria Harbour
 </td>
 <td align="center" width="50%">
 <img src="examples/london-westminster.png" alt="London Westminster editorial poster" width="100%"> 
 London · Westminster
 </td>
 </tr>
</table>

## Install

Clone the repository into your Codex skills directory:

```bash
git clone https://github.com/iiArius/liminal-editorial-posters.git ~/.codex/skills/liminal-editorial-posters
```

Restart or reload Codex after installation if the skill is not discovered immediately.

## Use

Upload one or more photographs, then invoke:

```text
$liminal-editorial-posters
```

Each image is processed separately. The skill requires an available image-generation or image-editing tool for final raster output.

## Design language

The core idea is **Liminal Frame**: a real subject or connected visual story crosses a quiet, low-saturation geometric window between documentary photography and editorial abstraction.

The skill deliberately distinguishes a single focal story from a single isolated object. For example, a city scene may preserve a tower, skyline, bridge, and river as one narrative ensemble.

## Repository structure

```text
liminal-editorial-posters/
├── SKILL.md
├── README.md
├── agents/
│   └── openai.yaml
└── examples/
    ├── hong-kong-victoria-harbour.png
    ├── london-westminster.png
    ├── san-francisco-golden-gate.png
    └── shanghai-pudong.png
```

## 中文简介

该 Skill 会把每张上传照片分别制作成一张独立的 3:4 高级编辑海报：上半部严格保留原始摄影，下半部提取主体或完整叙事组合，通过低饱和几何窗口、越界关系、东方留白与克制排版完成视觉重构。

liminal-editorial-posters

파일 탐색기

# CDN으로 사용하기

Was this content helpful?

Recent Feedback

Download README