Token导航 LogoToken导航

Agent Skill · 前端设计

poster-designer(海报设计师)

poster-designer 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

poster-designer 是前端设计类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为操作浏览器;许可证信息还不明确,第三方扫描已通过,按安装量在前端设计类的 10,216 个 Skill 里排第 3,272。

总安装 1,426周安装 60GitHub Stars 233下载 499核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:poster-designer(海报设计师)
来源仓库:https://github.com/xorbitsai/xagent
仓库路径:skills/poster-designer
安装命令:
npx skills add https://github.com/xorbitsai/xagent --skill poster-designer
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/xorbitsai/xagent --skill poster-designer

这个 Skill 是做什么的

前端设计类 Skill 关注页面结构、视觉规范和组件写法,输出通常是可运行的界面代码。

海报设计师是一种具有意图意识、遵守规则的海报生成技能。它保证使用确定性布局生成信息量大的海报,同时严格为最小的、仅视觉的用例保留直接图像生成。

每周安装量 60 存储库 xorbitsai/xagent GitHub 之星 233 第一次看到 今天 安全审计 Gen Agent Trust Hub 通行证 套接字通行证 斯尼克通行证

来源仓库内路径:skills/poster-designer

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/xorbitsai/xagent --skill poster-designer;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要操作浏览器,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

poster-designer 怎么安装?

在终端执行 npx skills add https://github.com/xorbitsai/xagent --skill poster-designer。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

poster-designer 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

poster-designer 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

poster-designer 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

poster-designer 用的人多吗?

公开数据是累计安装 1,426 次,近一周安装 60 次,下载 499 次,GitHub 233 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

poster-designer 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理前端设计类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

前端设计类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
frontend-design

前端设计 · Codex、Claude

8,823,360126,072只读MIT通过
vercel-composition-patterns

Vercel 组合模式 · Codex、Claude

3,858,81625,893需要联网MIT通过
vercel-react-native-skills

Vercel React native skills 前端 · Codex、Claude

2,468,66425,878只读MIT通过
adapt

adapt 搜索 · Codex、Claude

2,027,04023,125操作浏览器未明确通过
bolder

bolder 分析 · Codex、Claude

2,004,28823,130只读未明确通过
poster-designer

海报设计师 · 本页

1,426233操作浏览器未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文


Poster Designer

Overview

Poster Designer is a unified poster generation skill that always outputs one final poster image, while internally selecting the appropriate generation strategy.

Unlike soft, heuristic-based tools, this skill enforces hard routing rules to ensure that information-heavy posters are generated using deterministic layout rather than direct image generation.

The skill supports two internal strategies:

  • Layout-first (HTML-based) for posters that convey structured information
  • Image-first (direct image generation) for posters that are purely visual and symbolic

Strategy selection is final and non-negotiable once routing conditions are met.


When to Use

Use this skill whenever the desired output is:

  • A poster / long banner / vertical visual
  • A single image intended for publishing or long-term display

Typical use cases include:

  • Open-source milestone announcements
  • Community or project notices
  • Recruiting posters
  • Product or platform announcements
  • Technical infographics

Output Contract

This skill always produces:

  • Exactly one final poster image

The skill must never output:

  • HTML or CSS
  • Image-generation prompts
  • Design explanations
  • Intermediate artifacts

Only the final image is returned.


Strategy Selection (Hard Routing)

Poster Designer applies mandatory routing rules when selecting the generation strategy.

These rules override upstream planner preferences and must not be bypassed.


Mandatory Layout-first Conditions

The Layout-first (HTML-based) strategy MUST be used if any of the following conditions are satisfied:

  • The poster is described as an announcement, notice, or formal milestone
  • The poster is intended for long-term display, including but not limited to:

- README files - Official websites - Documentation pages - Community announcements

  • The request explicitly mentions or implies:

- structured information - clear hierarchy - readability - sections, paragraphs, or multiple text blocks

  • The poster contains more than one semantic text role, such as:

- a title - a numeric highlight - explanatory or descriptive text - acknowledgements or credits - footer or meta information

When any of these conditions are met:

  • Direct image generation MUST NOT be used
  • The task must be treated as a layout-driven artifact
  • HTML-based composition is required before producing the final image

Layout-first Strategy (HTML-based)

Characteristics

  • Fixed-width layout (default: 1080px)
  • Content-adaptive height (no hard-coded height)
  • Deterministic typography, spacing, and alignment
  • Clear reading order and visual hierarchy
  • Final poster image generated via full-container screenshot

Critical Screenshot Requirement

When using browser screenshot tools, the viewport width MUST be explicitly set to match the layout width (default: 1080px), and the output filename MUST clearly indicate the poster's purpose.

Intent

This strategy prioritizes:

  • Text clarity and legibility
  • Information hierarchy
  • Long-term usability and stability

It is the default strategy for any poster that functions as an information artifact rather than a purely visual symbol.


Image-first Strategy (Direct Generation)

Usage Constraints

The Image-first strategy is ONLY permitted when all of the following conditions are satisfied:

  • The poster contains a single headline or a single numeric highlight
  • Supporting text is minimal and decorative (no paragraphs or sections)
  • The request prioritizes atmosphere, emotion, or visual impact over information clarity
  • The poster resembles a keynote KV, launch visual, or symbolic milestone image

If any of these conditions are not met, Image-first generation must not be used.

Characteristics

  • Composition-driven design
  • Cinematic lighting and atmosphere
  • Text treated as a graphic element rather than structured content

Visual Design Principles

Regardless of strategy, Poster Designer enforces the following principles:

  • Structure before decoration
  • Clarity before effects
  • Semantic visuals over abstract decoration

The poster must clearly communicate:

  • What the poster represents
  • Why it matters
  • What the primary visual anchor is

Semantic Visual Anchors

Poster Designer introduces domain-specific visual anchors instead of generic backgrounds.

Examples include:

  • Docker milestones → containers, image layers, registry-to-node pull networks
  • Recruiting → structured sections, engineering environments, technical diagrams
  • Product platforms → architecture diagrams, UI-like compositions
  • Community milestones → nodes, constellations, global networks

Purely abstract backgrounds without semantic meaning are insufficient.


Typography & Text Handling

  • Clear hierarchy between headline, supporting text, and meta information
  • Text quantity adapts to the selected strategy
  • Text must remain legible at poster scale
  • Decorative effects must never reduce readability

Image Usage

  • Images are purposeful, not decorative fillers
  • Background imagery must preserve text readability
  • Mid-poster visuals must reinforce meaning
  • No competing focal points are allowed

Element Overlap & Visual Clarity Check

Before final screenshot, verify:

  1. No critical content is obscured

- Main title, key numbers, primary CTAs must be fully visible - Background imagery must not compete with text readability - Text placed over images should only be done when the obscured image area is non-essential

  1. Layer discipline

- Fixed-position elements must not cover center content - Decorative elements should not obscure semantic content - High z-index elements (tooltips, popups, overlays) should be user-triggered, not persistent

  1. Visual hierarchy

- Primary visual anchor should be immediately apparent - No competing focal points that confuse the viewer - Clear separation between foreground content and background decoration

  1. Text-on-image safety

- When text is placed over images, ensure sufficient contrast - Avoid placing text over areas with high visual detail that would reduce legibility - Solid or gradient overlays may be used to improve text readability

If any of these checks fail, the poster must be adjusted before taking the final screenshot.


Quality Bar

A successful output must be suitable for:

  • GitHub README banners
  • Docker Hub or open-source announcements
  • Conference or meetup visuals
  • Company or community social posts

If the output resembles:

  • A dashboard or analytics screenshot
  • A generic AI marketing image
  • An abstract background with loosely placed text

Then the skill has failed.


Example Invocation (Conceptual)

"Create a formal milestone announcement poster for Xinference celebrating 5,000,000 Docker Hub downloads. The poster must include a title, a prominent numeric highlight, explanatory text, acknowledgements, and footer information suitable for long-term display on README and official websites."


Summary

Poster Designer is an intent-aware, rule-enforced poster generation skill.

It guarantees that information-heavy posters are generated using deterministic layout, while reserving direct image generation strictly for minimal, visual-only use cases.

平台分布

Codex

37.68%
按下载量换算188

Claude

30.54%
按下载量换算152

Cursor

18.75%
按下载量换算94

Gemini CLI

8.92%
按下载量换算45

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills