Skip to content
Usable Docs

Fragment Types: Defaults, Custom Types, and How AI Uses Descriptions

Category: Getting Started | Difficulty: Beginner | Last Updated: 2025-10-21

Overview

Fragment types are organizational buckets that guide both humans and AI in classifying content.

What You’ll Learn

  • Default fragment types and when to use them
  • How to define custom types (e.g., Issues)
  • How AI uses type descriptions to store and retrieve content

Prerequisites

  • Basic understanding of fragments

Estimated Time

⏱️ 6 minutes

Concepts

Default Types

  • Knowledge: general documentation and reference
  • Recipe: step‑by‑step guides and tutorials
  • Solution: troubleshooting and fixes
  • Template: reusable code/config patterns
  • Instruction Set: rules and behaviors the AI considers first

Creating Custom Types

Example: “Issues” for bugs, feedback, questions (choose icon and color).

How AI Uses Descriptions

AI reads the fragment type description to decide where to store content, improving classification and retrieval.

Best Practices

  • Keep descriptions clear and action‑oriented
  • Avoid overlapping meanings across types
  • Memory Fragments: What They Are and How to Create One

Footer Note: This documentation is part of the Usable Public workspace and is automatically synced to the docs site.