Choose conventions that make your pages easier to write and find, and add structure when it solves a problem you actually have.
Start with top-level pages when you do not need a folder hierarchy. Links connect related subjects, and tags can place the same page in several categories. You can use folders when a hierarchy is useful: Folder/Page appears under Folder/ in the File Tree. Run Navigate: Tree to open it.
Common folder conventions include Inbox/ for quick notes and Journal/ for dated entries. These are useful patterns, not prerequisites for taking notes. Library/ is different: it holds installed libraries and customizations, and space-lua and space-style blocks only run from pages there. Guide/Finding Content compares the navigation options.
Use the visual Frontmatter editor to add page tags and attributes. Keep attribute names and types consistent across pages you want to query together. A standalone hashtag paragraph also tags the page.
If every page of a kind uses the same tag, you can query them together even when their other fields differ. A tag on a task categorizes that task. It does not tag the page. Guide/Add Structure and Query introduces page tags and attributes, and Markdown/Hashtags and Attribute retain the precise rules.
[[Folder/Page]]. The path does not depend on where the linking page lives. See Link.Begin in Workbench/Data to inspect records and build filters. Guide/Add Structure and Query teaches this without handwritten syntax. When you adapt the source:
index.contentPages() for ordinary page lists. It excludes tool pages such as library and configuration pages. Use index.pages() when you want those included.index.tasks() reads tasks. index.contentPages() reads pages. Both can be narrowed to a tag.Object Index owns the index helpers and user-visible behavior.
Reusable space-lua definitions are optional. Use Workbench/Console for small experiments before installing space-wide code on a page under Library/. When you change a definition, run System: Reload, then inspect Scripts and Console for errors. Reload completing does not establish that all scripts loaded correctly.
System: Reload calls API/system#system.reboot, which incorporates external changes to the open page and waits for indexing. Call it directly when a script needs that same readiness behavior. Put index-affecting tag definitions in ^Library/Config, and reindex existing content after changing a transform. Ordinary Markdown edits do not need this reload loop. Space Lua and Guide/When Results Update explain the distinct lifecycles.