Skip to main content
skill CX Codex

information-architecture · rampstackco — Claude Code skill: use and install

The information-architecture skill plans website or product structure: sitemaps, navigation, URLs, content types, and taxonomies. Use it for new IA work or a restructure.

MIT 354 installs ★ 1,107 updated 2026-10-19

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add rampstackco/claude-skills --skill information-architecture -a claude-code
Open on GitHub

What information-architecture does

The skill turns a product or website brief into an information architecture plan. It covers audience mental models, sitemap design, URL patterns, navigation, taxonomy, and labels. Its research methods include card sorting, tree testing, and first-click testing.

It also sets guidance for page hierarchies, breadcrumbs, categories, tags, metadata, and cross-page relationships. The approach is stack-agnostic. A useful result depends on having the scope, audience tasks, planned or existing content, and project constraints ready first.

What it helps with

  • map a new website’s page hierarchy
  • define stable URL and slug patterns
  • separate categories from tags and metadata
  • test whether users can find key content
  • audit and restructure an existing IA

Example prompts for the agent

  • Plan a sitemap for this product and mark each page type.
  • Review the URL structure for /products and /blog.
  • Design navigation, breadcrumbs, and a taxonomy for this knowledge base.

How to install information-architecture

1Skills CLI (Vercel)

npx skills add rampstackco/claude-skills --skill information-architecture -a claude-code

2Manually: put the files in place

~/.claude/skills/information-architecture/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/information-architecture/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about information-architecture

How do I install information-architecture in Claude Code?

Run npx skills add rampstackco/claude-skills --skill information-architecture -a claude-code, then open ~/.claude/skills/information-architecture/SKILL.md.

What does information-architecture do?

It plans the structure around a site or product, including the sitemap, URLs, navigation, content types, taxonomy, and labels. It can also guide an IA audit or restructure.

What does information-architecture need before it starts?

Prepare the product or site scope, the audience and their tasks, existing or planned content, and constraints such as parent IA, regulatory, or technical requirements.

Source text

SKILL.md · MIT
---
name: information-architecture
description: "Design the structure of a website or product including sitemap, navigation, URL structure, content types, taxonomy, and labeling. Use this skill whenever the user asks to plan a sitemap, design navigation, structure URLs, define content types, build taxonomies, design site search, or organize content at the system level. Triggers on sitemap, site structure, navigation, IA, information architecture, URL structure, content types, taxonomy, categorization, breadcrumbs, hub pages, faceted navigation, site search, labeling. Also triggers when content is being created without a structural plan, or when an existing site's structure is being audited or restructured."
category: strategy-and-discovery
catalog_summary: "Sitemap, navigation, URL structure, content types, taxonomy"
display_order: 4
---

# Information Architecture

Design the structure that holds the content. Stack-agnostic. Applies to marketing sites, product surfaces, knowledge bases, e-commerce, and editorial content.

A well-designed IA makes the rest of the project easier. A poorly-designed IA forces every downstream decision to fight the structure.

---

## When to use

- Designing a new site or major section from scratch
- Restructuring an existing site
- Adding a new content type or category
- Designing site navigation or menu systems
- Defining URL structure and slug patterns
- Building taxonomies or tag systems
- Auditing an existing IA for problems

## When NOT to use

- Single-page design (use `design-standards`)
- Content production (use `content-and-copy`)
- SEO-driven content planning (use `seo-keyword`)
- Initial brand and audience discovery (use `brand-discovery`)

---

## Required inputs

- The site or product scope
- The audience and what they're trying to do
- The content that exists or is planned
- Any constraints (parent IA, regulatory, technical)

If audience is unclear, run `brand-discovery` first. If content scope is unclear, run `content-strategy` first.

---

## The framework: 6 layers

Information architecture has six layers. Each builds on the one below.

### 1. Mental models

Before structure, understand how the audience thinks about the domain.

- What concepts do they group together naturally?
- What words do they use? (Often different from what the company uses.)
- What is the dominant frame of reference? (By task? By role? By topic? By time?)
- What do they expect to find where, based on conventions in similar products?

**Methods:**

- **Card sorting** (open or closed): Give the audience the content items, ask them to group them. Open card sorts surface natural groupings. Closed card sorts validate proposed groupings.
- **Tree testing:** Give a proposed structure, ask users to find specific items. Surfaces where the structure breaks down.
- **First-click testing:** Given a goal, where do users click first? If first clicks are wrong, the labels and structure are wrong.

### 2. Sitemap

The map of all pages and how they relate.

**Sitemap deliverables:**

- A hierarchy diagram showing parent-child relationships
- Indication of page types (static, dynamic, listing, detail)
- Cross-references showing how pages relate beyond the hierarchy
- Sometimes a separate user-flow overlay for key journeys

**Sitemap types:**

- **Hub-and-spoke** (cornerstone content + supporting content): Common for content marketing
- **Tree** (strict hierarchy, every page has one parent): Common for product documentation
- **Faceted** (content lives in many overlapping categories): Common for e-commerce
- **Flat** (everything reachable from the home): Common for small sites

Most sites blend types. Pick the dominant pattern and document the exceptions.

### 3. URL structure

URLs are part of the IA. They are user-facing, indexed by search engines, and shape how content is referenced.

**URL principles:**

- Reflect the content hierarchy
- Lowercase, hyphen-separated
- Predictable (same pattern across same content type)
- Stable (URLs don't change without redirects)
- Short (under 60 characters where possible)
- Descriptive (slug indicates the content)
- Free of dates unless time-bound
- Free of session IDs and tracking parameters in canonical form

**Common patterns:**

```
/                                   home
/[section]                          section landing
/[section]/[subsection]             subsection landing
/[section]/[subsection]/[item]      detail page
/blog                               blog index
/blog/[slug]                        blog post
/blog/category/[category]           category index
/blog/tag/[tag]                     tag index
/products                           product catalog
/products/[category]                category page
/products/[category]/[product]      product detail
```

Pick a pattern and stick to it. Inconsistent URL patterns confuse users, crawlers, and analytics.

### 4. Navigation

The chrome that gets users where they need to go.

**Primary navigation:**

- The top-level structure of the site
- Should reflect what the audience cares about, not what the org chart looks like
- 5 to 7 items maximum (more becomes cognitively heavy)
- Each label is recognizable in 2 to 3 words
- Order matters (left/first gets the most attention)

**Secondary navigation:**

- Within-section navigation
- Often shown as sidebars, sub-menus, or in-page tabs
- Supports the primary nav, doesn't duplicate it

**Utility navigation:**

- Account, search, login, support
- Visually subordinate to primary nav
- Often top-right (LTR languages)

**Breadcrumbs:**

- For nested hierarchies (3+ levels deep)
- Always linked except the current page
- Match the URL hierarchy or the conceptual hierarchy
- Marked up with BreadcrumbList schema

**Footer navigation:**

- Comprehensive; sometimes includes everything
- Organized by category for findability
- Includes secondary content (privacy, terms, contact)

### 5. Taxonomy and metadata

The classification system applied to content.

**Categories:**

- A small, controlled list (typically 5 to 15)
- Mutually exclusive ideal (one item, one category)
- Used for structural navigation

**Tags:**

- A larger, often growing list (50+)
- Multi-assignment (one item, many tags)
- Used for cross-cutting connections, related-content, and long-tail discovery

**Metadata fields:**

- Author, date, content type, audience segment
- Whatever is useful for filtering, sorting, and surfacing

**Common failures:**

- Categories that overlap (item could go in 3 different categories)
- Tags that are unmaintained (sprawl into thousands, become useless)
- Metadata fields that get filled inconsistently
- Different content types using different taxonomies for the same thing (chaos)

### 6. Labeling

What you call things.

**Label principles:**

- Audience language, not internal language
- Specific enough to be useful, short enough to scan
- Consistent across the site (call it "Product" or "Solutions" but not both)
- Tested with real users (closed card sort or tree test surfaces label problems)

**Common label problems:**

- "Solutions" (vague; usually means "products with marketing copy")
- "Resources" (catch-all; everything ends up there)
- Internal jargon ("PRD," "OKRs") that doesn't match user vocabulary
- Labels that change meaning across the site

---

## Workflow

1. **Understand the audience and content.** Use existing discovery and content strategy if available.
2. **Card sort or interview** to surface mental models.
3. **Draft the sitemap.** Hierarchy, page types, cross-references.
4. **Define URL patterns.** One pattern per content type.
5. **Design navigation.** Primary, secondary, utility, footer, breadcrumbs.
6. **Build taxonomy.** Categories (controlled, small) and tags (open, large).
7. **Validate labels.** Tree test or closed card sort with target users, or state the gap per the data-availability rule.
8. **Document.** Use the template in [`references/ia-document-template.md`](references/ia-document-template.md).
9. **Hand off to design and development.** IA decisions inform navigation components, URL routing, and taxonomy implementation.

---

## Failure patterns

- **IA designed by org chart.** "Engineering" and "Marketing" sections make sense to the company, not to the audience.
- **Categories that proliferate.** Every team adds a category for their thing. Becomes unscannable. Hold the line at 5 to 15.
- **Tags that sprawl.** No tag governance. Tags become a junk drawer.
- **Inconsistent URL patterns.** Some posts at /blog/[slug], some at /[slug], some at /articles/[slug]. Pick one.
- **Navigation that hides primary content.** The most important pages should be one click from home.
- **Search as a substitute for IA.** "Just use search" is not a structure. Search supports IA, doesn't replace it.
- **No validation.** Card sorts, tree tests, and first-click tests are cheap and surface huge problems early.
- **Treating IA as a one-time deliverable.** IA evolves with content. Plan for evolution.

---

## Output format

Default output is an IA document at `information-architecture.md` plus visual assets:

1. Executive summary
2. Audience and mental models (synthesized)
3. Sitemap (hierarchical diagram)
4. URL structure (per content type)
5. Navigation specification (primary, secondary, utility, footer, breadcrumbs)
6. Taxonomy (categories and tag governance)
7. Labels (validated wording for navigation, categories, content types)
8. Implementation notes for design and development

Visual deliverables:
- Sitemap diagram (Whimsical, Figma, OmniGraffle, etc.)
- Navigation wireframes for primary surfaces
- Optional: card sort and tree test results

---

## If required data is unavailable

This skill's output depends on data, measurements, or tool results it cannot generate on its own. When a required input, tool, or data source is unavailable or unverifiable, the sanctioned output is the deliverable with the gap stated: what was needed, what was actually obtained or verified, and which parts of the output are affected. Fabricating, estimating, or interpolating a required number to complete the deliverable is never sanctioned. A stated gap is a complete answer.

---

## Reference files

- [`references/ia-document-template.md`](references/ia-document-template.md) - Template for the IA deliverable.
- [`references/url-pattern-library.md`](references/url-pattern-library.md) - URL pattern conventions for common content types.

© rampstackco · MIT · rampstackco/claude-skills

Similar skills

All skills for Design →

information-architecture by other authors comes first.

information-architecture julianoczkowski information-architecture defines product structure and navigation before visual design begins. Bridges the gap between a design brief and actual frontend code. CX Codex +1 6,300installs information-architecture Owl-Listener information-architecture is a Claude Code skill for organizing existing content into categories, labels, and navigation. It activates when a product’s sections and materials need a clearer structure. CX Codex +1 2,400installs information-architecture dembrandt information-architecture keeps a large app navigable: what to call each entity, how screens map to the data model, and when to ask before a destructive action. CX Codex +1 1,100installs information-architecture aj-geddes information-architecture structures content, sitemaps, and taxonomies for web apps and documentation hubs. It guides your AI agent through a four-stage process from card sorting to tree testing. CX Codex +1 642installs information-architecture lyndonkl information-architecture — a skill for organizing content so users find what they need. Guides through audits, card sorting, taxonomy design, navigation trees, and validation testing. CX Codex +1 127installs information-architecture-navigation hueyexe information-architecture-navigation — a skill for designing navigation, labels, taxonomy and search in a digital product. Useful when the structure of a site or app stops helping people find things. CX Codex +1 664installs information-architecture-and-navigation podo Guides your agent through information architecture and navigation systems using the polar bear model. Useful when organizing taxonomies, site structures, and findability tests. CX Codex +1 222installs organize ghaida organize designs information architecture, navigation models, and taxonomy systems. Helps when users get lost or struggle to locate features. CX Codex +1 3,400installs site-architecture coreyhaines31 site-architecture — a skill for planning page hierarchy, URL structure, navigation, and internal linking when building or restructuring a website. Works on Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, and OpenCode. CX Codex +1 135,700installs information-architecture-reviewer ShuaiMXu information-architecture-reviewer is a Claude Code skill that audits navigation and content hierarchy against user tasks. It identifies dead ends and ambiguous labels along target flows. CX Codex +1 ★ 529repo stars critique-information-density Owl-Listener critique-information-density — a skill that critiques a rendered screen's density: cognitive load, content priority, scanning patterns and progressive disclosure. Useful when a screen feels overwhelming. CX Codex +1 2,200installs

From the same set102

The whole set →
design-system Builds or audits a design system: tokens, component library, naming conventions, contribution model and docs. Useful when products drift apart visually. CX Codex +1 369installs brand-voice Builds a complete brand voice and tone guide with personality traits, contextual shifts, vocabulary rules, and before-and-after copy pairs. CX Codex +1 367installs brand-discovery brand-discovery is a Claude Code skill that researches audiences, competitors, and category dynamics before a brand or website project. It helps explore positioning options before a creative brief. CX Codex +1 366installs accessibility-audit accessibility-audit audits web interfaces against WCAG 2.1 AA standards and builds a prioritized remediation plan. Helps prepare sites for ADA, Section 508, or routine health checks. CX Codex +1 364installs creative-brief creative-brief turns a vague idea into a concrete brief that a designer or developer can pick up and run with. Use it to lock down scope, audience, and success metrics before the first line of code. CX Codex +1 354installs seo-technical seo-technical runs a technical SEO audit covering crawlability, indexability, rendering, site architecture, structured data, Core Web Vitals, security and internationalization. CX Codex +1 347installs seo-onpage seo-onpage runs one page through eight on-page checks: title, meta description, headings, copy, internal links, images, slug and schema. Returns a prioritised report. CX Codex +1 347installs quiz-and-assessment-design quiz-and-assessment-design is a Claude Code skill that plans quizzes and assessments, linking outcomes to recommendations. It supports lead magnets and diagnostics. CX Codex +1 345installs

Project sponsors

Companies that keep this analytics open