Concepts
Knowledgebases are versioned collections of pages and snippets.
Within a workspace, each knowledgebase is the unit of authored knowledge. It contains pages, snippets, embeds, revisions, and releases. This is the level where teams structure content for reuse and publishing.
Model
The canonical objects inside a knowledgebase
Snippets
Embeds and blocks
Revisions
Releases
Environment pointers
Behavior
How a SnippetGraph knowledgebase behaves differently
The main difference is that knowledge is modeled as connected objects instead of isolated files. If a snippet is reused in twenty-seven pages, the system knows that. If a publish would affect three export targets, the system knows that too.
That visibility changes editing behavior. Teams no longer guess what a shared change might break. They can inspect impact, review composed output, and publish with confidence.
Best practice
How to structure a strong knowledgebase
Keep snippets narrow enough to be reusable, but not so tiny that pages become unreadable mosaics. A good snippet usually represents a stable policy fragment, procedural sequence, or reusable answer block.
Keep pages focused on a user journey or operational task. The page should read naturally as a document, even though some of its content is composed from shared modules underneath.
Next move
Model shared truth once, then compose it into the pages people actually use.
That is the design center of the product. Pages stay readable. Snippets stay governable. Releases stay predictable.