How I Organize My Second Brain Using Obsidian Bases + Claude Code

The core shift is away from folder-based “where does this note belong?” decisions and toward a tag-and-database architecture where notes live in one place, then get retrieved through dynamic views. The payoff is less capture friction and faster retrieval: instead of spending time choosing folders and later hunting through subfolders, users create a note, apply a template that auto-fills metadata, and then navigate by “what it is” (categories) and “what it’s about” (subjects). That change matters because folder systems impose a hidden maintenance cost—every time a second brain is opened, the user must manage structure rather than think.

The system starts with five top-level folders in Obsidian: an inbox for unprocessed captures, a single flat notes folder where every note is stored with no subfolders, two navigation layers—categories and subjects—and a system folder for templates, attachments, and dashboards. The philosophy is explicit: folders decide location in traditional setups, but this design decides meaning through metadata. Every note is tagged with category and subject properties, and Obsidian’s built-in “basis” (Obsidian Bases) turns those links into database-style container views.

Categories function like filterable containers for different object types—permanent notes, newsletters, YouTube scripts, tweets, SOPs, AI prompts, projects, books, and more. Each category is implemented as a container note that embeds a Basis database view. The view’s rule is simple: it shows every note that links to that category. YAML front matter at the top of each note stores these links, so when a note includes a category property (e.g., “YouTube videos”), it instantly appears inside the corresponding container without dragging, sorting, or manual indexing. Unlike folder placement, a single note can belong to multiple categories at once, enabling “find from any angle” retrieval.

Subjects add a second navigation layer that cuts across categories. If categories answer “what type of note is this?”, subjects answer “what is it about?”—examples include business, creativity, psychology, philosophy, health, productivity, and relationships. A newsletter can simultaneously sit in the newsletter category container and the psychology subject container, so users can browse by object type or by theme without relying on invisible tag piles.

Capture becomes streamlined through templates tied to categories. When creating a new note, it lands in the notes folder automatically. Then a category-specific template applies a pre-written YAML block that fills properties like category links, status, and type. The user’s job is reduced to naming and writing; the system handles organization. Retrieval is then done via Obsidian’s quick switcher by category or subject, matching how people naturally think (“I need a newsletter” or “I need AI notes”) rather than how folder systems force users to reason (“project vs resource vs area”).

Maintenance is delegated to Claude Code running inside the vault. Claude Code reads cloud.md files in each folder to understand the vault architecture and the meaning of categories, then can create and update notes automatically. Examples include generating a newsletter note from a prompt while pulling prior newsletters for voice and tone, processing the inbox by assigning correct categories and subjects, and bulk-updating metadata across many notes. The result is a compounding second brain: knowledge captured today becomes connected context for content created later.

For convenience, a pre-built vault called “Sovereign Creator OS Light” is offered as a ready-to-open Obsidian vault with preconfigured category and subject containers, templates, Claude Code prompts/skills, SOPs, and cloud.md files so the system works quickly after setup. Beyond organization, the stated goal is long-term compounding—turning the vault into a foundation for a creator business and broader life fulfillment.