ia

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Information Architecture Generator

信息架构生成工具

Generate a complete Information Architecture (IA) for a product based on a service plan, PRD, or startup idea.
根据服务方案、PRD(产品需求文档)或创业想法,为产品生成完整的信息架构(IA)。

When to Use This Skill

何时使用该Skill

Use this skill when the user provides:
  • A service planning document
  • A PRD (Product Requirements Document)
  • A startup or product idea description
Trigger phrases: "IA 작성", "정보구조도 만들어줘", "화면 위계 정리해줘", "create IA", "generate IA", "make information architecture"

当用户提供以下内容时,可使用本Skill:
  • 服务规划文档
  • PRD(产品需求文档)
  • 创业或产品想法描述
触发短语:"IA 작성"、"정보구조도 만들어줘"、"화면 위계 정리해줘"、"create IA"、"generate IA"、"make information architecture"

Knowledge Base: Navigation Flow Design Principles

知识库:导航流程设计原则

An app or service is fundamentally the automation of user data manipulation. The goal of screen design: translate the user's mental model and task flow into DB operations (CRUD).
Scope: This skill generates frontend routes (browser URLs). API endpoint design is a separate concern.
应用或服务从本质上来说是用户数据操作的自动化。 屏幕设计的目标:将用户的心智模型和任务流程转化为数据库操作(CRUD)
范围说明:本Skill生成前端路由(浏览器URL)。API端点设计属于独立范畴。

Step 1: Define Entities

步骤1:定义实体

Identify the core concepts (entities) the service handles, then map each user action to a CRUD operation.
User ActionDB OperationExample (E-commerce)
CreateCreateSign up, add product, place order
ViewReadView profile, search/view product, order history
EditUpdateChange password, update price, change order status
RemoveDeleteDelete account, remove product
识别服务处理的核心概念(实体),然后将每个用户操作映射到CRUD操作。
用户操作数据库操作示例(电商场景)
创建Create注册、添加商品、下单
查看Read查看个人资料、搜索/查看商品、订单历史
编辑Update修改密码、更新价格、变更订单状态
删除Delete删除账户、移除商品

Step 2: Design Screen Hierarchy

步骤2:设计屏幕层级

Naming Rules:
  1. Plural nouns for collections
    /posts
    ,
    /users
    ,
    /fields
    (not
    /post
    ,
    /field
    )
  2. Entity-based naming — use data model names (
    /tasks
    ), not UI labels (
    /warehouse
    ,
    /feed
    )
  3. Minimize action segments — when a separate page is unavoidable, use a fixed set:
    create
    ,
    edit
    ,
    settings
1-Depth: Top-Level — Entity Collections
  • URL pattern:
    /{entities}
    (e.g.
    /posts
    ,
    /users
    ,
    /tasks
    )
  • Non-resource pages:
    /{static-page}
    (e.g.
    /about
    ,
    /pricing
    )
2-Depth: Record Detail & Entry Points
  • Rule: A single page handles only one entity unless absolutely necessary
  • /{entities}/:id
    — Dynamic route, specific resource detail (e.g.
    /posts/123
    )
  • /{entities}/create
    — Creation form (prefer modal when the form is simple)
3-Depth: Sub-pages of a Parent Record
  • /{entities}/:id/{sub-entities}
    — Parent owns child records (e.g.
    /farms/5/members
    )
  • /{entities}/:id/settings
    — Configuration page for the parent
  • /{entities}/:id/edit
    — Edit form that needs its own page
  • Modal: simple action; preserve current page context
  • Inline: page handles view + edit together (e.g. settings with save button)
  • Separate page: multi-step form or independent flow
  • Modals and inline editing reduce the need for action routes
Query Parameters: Reflect Page State in URL
  • URL pattern:
    ...?key=value
  • Filtering (
    ?category=tops
    ), sorting (
    ?sort=price
    ), pagination (
    ?page=2
    ), search (
    ?q=keyword
    )
  • Status view (
    ?status=backlog
    ), UI mode (
    ?mode=edit
    )
  • Use query params for view state instead of separate routes (e.g.
    ?status=backlog
    not
    /warehouse
    )
命名规则
  1. 集合使用复数名词
    /posts
    /users
    /fields
    (而非
    /post
    /field
  2. 基于实体命名 — 使用数据模型名称(如
    /tasks
    ),而非UI标签(如
    /warehouse
    /feed
  3. 最小化操作分段 — 当必须使用独立页面时,使用固定的操作词:
    create
    edit
    settings
1级深度:顶层——实体集合
  • URL格式:
    /{entities}
    (例如
    /posts
    /users
    /tasks
  • 非资源页面:
    /{static-page}
    (例如
    /about
    /pricing
2级深度:记录详情与入口
  • 规则:单个页面仅处理一个实体,除非绝对必要
  • /{entities}/:id
    — 动态路由,特定资源详情(例如
    /posts/123
  • /{entities}/create
    — 创建表单(表单简单时优先使用模态框)
3级深度:父记录的子页面
  • /{entities}/:id/{sub-entities}
    — 父记录包含子记录(例如
    /farms/5/members
  • /{entities}/:id/settings
    — 父记录的配置页面
  • /{entities}/:id/edit
    — 需独立页面的编辑表单
  • 模态框(Modal):用于简单操作;保留当前页面上下文
  • 内联(Inline):页面同时处理查看与编辑(例如带保存按钮的设置页面)
  • 独立页面:多步骤表单或独立流程
  • 模态框和内联编辑可减少操作路由的使用
查询参数:在URL中反映页面状态
  • URL格式:
    ...?key=value
  • 筛选(
    ?category=tops
    )、排序(
    ?sort=price
    )、分页(
    ?page=2
    )、搜索(
    ?q=keyword
  • 状态视图(
    ?status=backlog
    )、UI模式(
    ?mode=edit
  • 使用查询参数表示视图状态,而非独立路由(例如使用
    ?status=backlog
    而非
    /warehouse

Step 3: Design Page Structure

步骤3:设计页面结构

Global Layout (areas shared across most pages):
  • Header: brand, logged-in user info, notifications
  • Sidebar: navigation aid / feature drawer (desktop & dashboards)
  • Footer: legal info, copyright, support, sitemap
  • Global Navigation: branches top-level paths; mobile → bottom tab bar or drawer
Content Area Types:
  • Collection: a set of records for one entity — list, table, menu (e.g.
    /posts
    )
  • Single Item: one element from a collection — detail view (e.g.
    /posts/123
    )
  • Sub-page: child records, settings, or edit form of a parent (e.g.
    /posts/123/settings
    )

全局布局(大多数页面共享的区域):
  • 页眉(Header):品牌标识、登录用户信息、通知
  • 侧边栏(Sidebar):导航辅助/功能抽屉(桌面端与仪表盘场景)
  • 页脚(Footer):法律信息、版权、支持、站点地图
  • 全局导航(Global Navigation):关联顶层路径;移动端→底部标签栏或抽屉菜单
内容区域类型
  • 集合:单个实体的一组记录——列表、表格、菜单(例如
    /posts
  • 单个条目:集合中的一个元素——详情视图(例如
    /posts/123
  • 子页面:父记录的子记录、设置或编辑表单(例如
    /posts/123/settings

How to Generate the IA

如何生成IA

Step 1: Analyze the Input

步骤1:分析输入内容

Read the provided service plan, PRD, or product idea. Extract:
  1. Core entities (what data does this service manage?)
  2. User roles (who uses this service? any admin/user separation?)
  3. Key features per entity (what CRUD operations are needed?)
  4. Navigation structure (what are the top-level sections?)
阅读提供的服务方案、PRD或产品想法,提取以下信息:
  1. 核心实体(该服务管理哪些数据?)
  2. 用户角色(谁使用该服务?是否区分管理员/普通用户?)
  3. 各实体的关键功能(需要哪些CRUD操作?)
  4. 导航结构(顶层板块有哪些?)

Step 2: Map Entities to CRUD

步骤2:将实体映射到CRUD操作

For each entity, list the screens required:
  • Collection screen (list of all records)
  • Single Item screen (detail of one record)
  • Action screens (create, edit, delete flows)
针对每个实体,列出所需的屏幕:
  • 集合屏幕(所有记录的列表)
  • 单个条目屏幕(单条记录的详情)
  • 操作屏幕(创建、编辑、删除流程)

Step 3: Build the Nested List

步骤3:构建嵌套列表

Output the IA as a Markdown nested list following this structure:
- /{entities}
  - Collection Screen (`/{entities}`)
    - Query params: `?status=`, `?sort=`, `?q=`
  - Single Item Screen (`/{entities}/:id`)
    - Sub-page (`/{entities}/:id/settings`) (modal, inline, or page)
    - Sub-page (`/{entities}/:id/edit`) (modal or page)
    - Child records (`/{entities}/:id/{sub-entities}`)
  - Create Screen (`/{entities}/create`) (modal or page)
- /{static-page}
Rules:
  • 1-depth items = top-level navigation sections (plural entity names or standalone pages)
  • 2-depth items = main screens within each section
  • 3-depth items = sub-pages, child records, or action forms of a parent
  • Resource paths use plural nouns and entity-based naming
  • Allowed action segments:
    create
    ,
    edit
    ,
    settings
    — prefer modals/inline to reduce these
  • Mark presentation format: append
    (modal)
    ,
    (inline)
    , or
    (page)
    where relevant
  • Mark dynamic routes: use
    /:id
    for record-specific screens
  • Use query parameters for view state (filtering, status) instead of separate routes
  • Include Global Layout section at the top (Header, GNB, Footer)
  • Include Auth flow if the service requires login
按照以下结构,以Markdown嵌套列表形式输出IA:
- /{entities}
  - 集合屏幕 (`/{entities}`)
    - 查询参数: `?status=`, `?sort=`, `?q=`
  - 单个条目屏幕 (`/{entities}/:id`)
    - 子页面 (`/{entities}/:id/settings`)(模态框、内联或独立页面)
    - 子页面 (`/{entities}/:id/edit`)(模态框或独立页面)
    - 子记录 (`/{entities}/:id/{sub-entities}`)
  - 创建屏幕 (`/{entities}/create`)(模态框或独立页面)
- /{static-page}
规则
  • 1级深度条目 = 顶层导航板块(复数实体名称或独立页面)
  • 2级深度条目 = 各板块内的主屏幕
  • 3级深度条目 = 父页面的子页面、子记录或操作表单
  • 资源路径使用复数名词基于实体的命名
  • 允许使用的操作分段:
    create
    edit
    settings
    — 优先使用模态框/内联方式以减少此类路由
  • 标注呈现格式:在相关位置追加
    (modal)
    (inline)
    (page)
  • 标注动态路由:对记录专属屏幕使用
    /:id
  • 使用查询参数表示视图状态(筛选、状态)而非独立路由
  • 在顶部包含全局布局板块(页眉、全局导航栏、页脚)
  • 如果服务需要登录,包含认证流程

Step 4: Add Screen Metadata

步骤4:添加屏幕元数据

After the nested list, include a table summarizing each screen:
ScreenTypeURL PatternDescription
...Collection / Single Item / Action / Auth
/...
...

在嵌套列表之后,添加一个表格汇总每个屏幕的信息:
屏幕类型URL格式描述
...集合 / 单个条目 / 操作 / 认证
/...
...

Output Format

输出格式

Always output:
  1. Service Summary — 1–2 sentences summarizing the product
  2. Entities & CRUD Map — bullet list of entities with their operations
  3. Information Architecture — full nested list
  4. Screen Summary Table — metadata table

始终输出以下内容:
  1. 服务摘要 — 1-2句话总结产品
  2. 实体与CRUD映射表 — 实体及其操作的项目符号列表
  3. 信息架构 — 完整的嵌套列表
  4. 屏幕摘要表格 — 元数据表格

Save Instructions

保存说明

When the task is complete, save the output to
SCREENS.md
in the current working directory unless the user specifies a different filename or location.
Use this format for the saved file:
markdown
undefined
任务完成后,将输出内容保存至当前工作目录下的
SCREENS.md
,除非用户指定其他文件名或位置。
保存文件时使用以下格式:
markdown
undefined

[Service Name] — Information Architecture

[服务名称] — 信息架构

Generated: [date]
生成时间: [日期]

Entities & CRUD Map

实体与CRUD映射表

...
...

Screen Hierarchy

屏幕层级

...
...

Screen Summary

屏幕摘要

...
undefined
...
undefined