Skip to content
Back to skills

Portrait 916

ASecurity

人像生图默认 9:16 竖幅规则。TRIGGER: 人像生图、portrait generation、人像、portrait、人脸、9:16、竖幅人像、韩系人像、网红人像

  • 54 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 22, 2026
ai-agentspython

Works with

  • cli

Security analysis

A100/100

Scanned September 22, 2026

npx -y skills add rexleimo/aios --skill portrait-916 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Portrait 916?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Portrait 916
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/rexleimo-portrait-916/badge)](https://www.skillsdirectory.com/skills/rexleimo-portrait-916)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: portrait-916
description: '人像生图默认 9:16 竖幅规则。TRIGGER: 人像生图、portrait generation、人像、portrait、人脸、9:16、竖幅人像、韩系人像、网红人像'

installCatalogName: portrait-916
clients: [codex, claude, gemini, opencode, hermes, workbuddy, pi, zcode, qoder]
scopes: [global, project]
defaultInstall:
  global: false
  project: false
tags: [portrait, image, generation, 9x16]
repoTargets: [codex, claude, gemini, opencode, hermes, workbuddy, pi, zcode, qoder]
---

# Portrait 9:16 — 人像生图画幅升格规则

生成人像(portrait)图片时,**默认一律用 9:16 竖幅**。这是实测结论,不是偏好。

## 核心规则

1. **人像 → 9:16**:`gpt-image-2` 用 `--size 1024x1792`。不要用 3:4 直接出人像。
2. **为什么**:3:4 画幅下模型把人脸摊进更宽的画面,五官更扁、皮肤纹理更弱、"抓拍感"更差;9:16 对应手机竖屏自拍的真实取景,人脸占比和景深逻辑都更对。2026-09-12 kr-realism 任务实测:同风格下 3:4 出图(gen/probe 时期)人脸明显弱于 9:16 v4(见 rex-videos `.workspace/2026-09-12-kr-realism/` 对比记录)。
3. **下游要 3:4 时**:先 9:16 生成,再用 Python(Pillow)裁成 3:4,禁止直接 3:4 生成后凑合。
4. **提示词里同步声明画幅**:prompt 首句写 `9:16 vertical ...`,与 `--size 1024x1792` 保持一致。

## 尺寸对照(gpt-image-2 支持值)

| 画幅 | size | 用途 |
|------|------|------|
| 9:16 | `1024x1792` | **人像默认** |
| 1:1 | `1024x1024` | 头像/方形位、快速试 prompt |
| 16:9 | `1792x1024` | 横版场景/背景图 |

## 执行要点

- 执行器用 `rexai-image-generation` 技能的 **Python 版** `rexai-image.py`(进度可见、超时 600s、熔断/内容策略自动重试)。
- 9:16 大尺寸出图常需 60-180s,属正常,勿中途 kill(job 在服务端继续跑,kill 会丢 job id 上下文)。
- 内容安全过滤是概率性的:同一 prompt 可能过可能拦;被拦时脚本自动重试,仍失败就替换高危措辞(实测 "one hand ... in foreground" 易拦,"leaning forward with hands on the bed edge" 稳定过)。

## 验证记录

- 2026-09-12:kr-realism 任务,ref-01 参考图拆解 → v4 prompt(9:16, 1024x1792)一次通过,8 项场景锚点全中,人脸立体度显著优于 3:4 对照图。

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…