AgentSkillsCN

building-taxonomy-and-ia

为内容组织构建信息架构与分类体系。在设计导航、分类系统，或构建内容层级结构时使用。

SKILL.md
--- frontmatter
name: building-taxonomy-and-ia
description: Creates information architecture and taxonomy structures for content organization. Use when designing navigation, categorization systems, or content hierarchies.

Building Taxonomy and Information Architecture

Quick start

Collect or infer:

  • •Content inventory (what exists or will exist)
  • •User mental models (how users think about the domain)
  • •Primary user tasks and goals
  • •Business categorization requirements
  • •Existing navigation or organization (if any)

Then produce output using TEMPLATES.md. Validate with RUBRIC.md.

Workflow

  1. •Inventory existing content and identify content types
  2. •Conduct card sorting or analyze search/navigation data for user mental models
  3. •Define taxonomy principles (mutual exclusivity, exhaustiveness, etc.)
  4. •Create hierarchical structure with max 3 levels of depth
  5. •Define naming conventions for categories
  6. •Map content types to taxonomy nodes
  7. •Test with representative user tasks
  8. •Run the rubric check. Revise until it passes.

Degrees of freedom

Freedom level: Low

  • •Default: follow templates exactly
  • •Allowed variation: depth of hierarchy (2-3 levels), number of top-level categories—as long as rubric passes
  • •Strict constraints: taxonomy must be mutually exclusive; no orphan content; depth cannot exceed 3 levels

State awareness

  • •Greenfield: Design IA before content creation; establish patterns early
  • •Existing chaos: Audit first, identify clusters, then rationalize
  • •Scaling content: Focus on scalable categories, not content-specific buckets
  • •Multi-product: Define shared taxonomy layer, product-specific branches

References