Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问许可证需确认审计通过

design设计

Agent Skill

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

总安装

648

周安装

27

GitHub Stars

318

下载量

216
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

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:design(设计)
来源仓库:https://github.com/boshu2/agentops
仓库路径:skills/design
安装命令:
npx skills add https://github.com/boshu2/agentops --skill design
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/boshu2/agentops --skill design

简介

design 验证目标是否符合 PRODUCT.md 战略方向,作为重大开发前门禁检查。

  • 支持快速模式与严格阈值设置,输出对齐度评分与偏离风险提示。
  • 若缺少 PRODUCT.md 则警告并引导创建,确保产品一致性。
  • 推荐与 /brainstorm 联用,避免资源浪费在非优先方向上。
  • design 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

/design -- Product Validation Gate

Quick Ref: Validates that a proposed goal aligns with the product's strategic direction before discovery begins major work. Council-gated with --preset=product.

Quick Start

/design "add caching layer for CLI"           # validate goal against PRODUCT.md
/design --quick "add caching layer"           # inline check, no council spawning
/design --strict "redesign hook system"       # higher threshold (avg >= 2.5)
/design                                       # infers goal from recent context

Execution Steps

Step 0: Check for PRODUCT.md

Locate PRODUCT.md in the repo root.

ls PRODUCT.md 2>/dev/null

If absent: Output a warning and return PASS with note: "No PRODUCT.md found -- skipping product validation gate. Run /product to generate one."

If present: Continue to Step 1.

Step 1: Load Product Context

Read PRODUCT.md and extract:

  • Mission -- the product's core purpose
  • Personas -- defined user types and their needs
  • Gaps -- known product gaps or roadmap items
  • Competitive landscape -- how the product differentiates

If any section is missing, note it as unavailable and score that dimension conservatively (1).

Step 2: Score Alignment Matrix

Evaluate the proposed goal against five dimensions. Use the scoring rubric in references/alignment-matrix.md.

DimensionScore (0-3)Rationale
Gap AlignmentDoes this goal address a known product gap?
Persona FitDoes this serve defined personas?
Competitive DiffDoes this strengthen competitive position?
PrecedentHas similar work been done before? What can we learn?
Scope FitIs this appropriately scoped for the current phase?

Compute the average score across all five dimensions.

Step 3: Run Council

Invoke council with the product preset. See references/product-council-preset.md for judge configuration.

Skill(skill="council", args="--preset=product validate design alignment for: <goal>")

Pass the alignment matrix from Step 2 as context to the council judges.

If --quick flag is set, skip council spawning and perform inline assessment instead.

Step 4: Write Design Artifact

Write the design artifact to .agents/design/:

mkdir -p .agents/design

Filename: <date>-design-<goal-slug>.md (e.g., 2026-03-30-design-add-caching-layer.md)

Artifact contents:

  • Goal statement
  • Alignment matrix with scores and rationale
  • Council verdict (or inline assessment if --quick)
  • Final verdict and recommendation

Step 5: Output Verdict

Determine the final verdict based on scores:

ConditionVerdict
Average score >= 2.0 AND no dimension at 0PASS -- goal aligns with product direction
Average score >= 1.5 OR one dimension at 0 with others strongWARN -- goal has alignment concerns, review before proceeding
Average score < 1.5 OR multiple dimensions at 0FAIL -- goal does not align with product direction

When --strict is set, raise the PASS threshold to average >= 2.5.

Output the verdict in this format:

DESIGN VERDICT: <PASS|WARN|FAIL>
  Gap Alignment:     <score>/3 -- <one-line rationale>
  Persona Fit:       <score>/3 -- <one-line rationale>
  Competitive Diff:  <score>/3 -- <one-line rationale>
  Precedent:         <score>/3 -- <one-line rationale>
  Scope Fit:         <score>/3 -- <one-line rationale>
  Average:           <avg>/3.0
  Artifact:          .agents/design/<filename>.md

Flags

FlagEffect
--quickInline check without council spawning. Faster but less thorough.
--strictRaise PASS threshold from avg >= 2.0 to avg >= 2.5. Use for high-stakes changes.

Reference Documents


See Also

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

35.16%
按下载量换算76

Claude

30.09%
按下载量换算65

Cursor

22.58%
按下载量换算49

Gemini CLI

10.15%
按下载量换算22

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills