Vault folders are created during /onboarding.

onebrain/
├── 00-inbox/          Raw braindumps and captures (process regularly)
│   └── imports/       Staging area for /import (drop files here)
├── 01-projects/       Active projects with inline tasks
├── 02-areas/          Ongoing responsibilities (health, finances, career...)
├── 03-knowledge/      Your own synthesized thinking and insights
├── 04-resources/      External info — research output, summaries, reference
├── 05-agent/          AI-specific context and memory
│   ├── MEMORY.md      Identity + Active Projects + Critical Behaviors
│   ├── INDEX.md       Index of all memory files (loaded every session)
│   └── memory/        All memory files — behavioral patterns, domain context, project facts
├── 06-archive/        Completed projects and archived areas
├── 07-logs/           Session logs and checkpoints (YYYY/MM/ subfolders)
├── attachments/       Copied files from /import --attach
│   ├── pdf/
│   ├── images/
│   └── video/
├── TASKS.md           Live task dashboard (created by /tasks, opened in Obsidian)
├── MOC.md             Vault portal — Map of Content (created by /moc)
├── CLAUDE.md          Instructions for Claude Code
├── GEMINI.md          Instructions for Gemini CLI
├── AGENTS.md          Universal agent instructions
├── vault.yml          Your vault configuration (created during onboarding)
└── .claude/plugins/   AI skills and hooks

The core workflow: capture everything to inbox → process with /consolidate → synthesize into knowledge or save as reference → archive what's done.

00-inbox/ — Raw braindumps and captures Process regularly. Everything unclassified lands here first. The imports/ subfolder is the staging area for /import — copy files there and run /import to distill them into vault notes.

01-projects/ — Active work with a clear goal and end date Examples: work/Website Redesign.md, personal/Japan Trip 2026.md

02-areas/ — Ongoing responsibilities that never "complete" Examples: health/Running Log.md, finances/Budget 2026.md

03-knowledge/ — Your own synthesized thinking Conclusions, frameworks, and insights you've developed — not raw reference material. Examples: productivity/Deep Work Principles.md, technology/When to Use Microservices.md

04-resources/ — External information saved for reference Output from /research, /summarize, /reading-notes, /import, and saved reference material. Examples: research/Zettelkasten Method.md, code-snippets/Go HTTP Middleware.md

05-agent/ — Your agent's portable mind Everything the AI knows about you. Copy this folder to move your agent to a new vault.

• MEMORY.md — Identity + Active Projects + Critical Behaviors — loaded every session
• INDEX.md — Index of all memory files — loaded every session alongside MEMORY.md
• memory/ — All memory files — behavioral patterns, domain context, project facts

06-archive/ — Completed projects and retired areas Organized by date archived: 06-archive/YYYY/MM/.

07-logs/ — Session logs and checkpoints Session logs: 07-logs/YYYY/MM/YYYY-MM-DD-session-NN.md — generated by /wrapup or auto-saved at session end. Checkpoints: 07-logs/YYYY/MM/YYYY-MM-DD-{PPID}-checkpoint-NN.md — auto-generated by hooks every 15 messages or 30 minutes, and before context compression. Incorporated and deleted by /wrapup when wrapping up.

OneBrain uses a four-tier memory system, where knowledge flows upward as it gets validated:

Tier 1 — Working memory (00-inbox/ + current session) Everything that hasn't been processed yet. Captures from /braindump, /capture, and quick notes land here. Process with /consolidate to move into the knowledge base.

Tier 2 — Episodic memory (07-logs/) Session logs: YYYY-MM-DD-session-NN.md in YYYY/MM/ subfolders. Contains summaries, decisions, insights, and action items from each session. Generated by /wrapup. Checkpoints: YYYY-MM-DD-{PPID}-checkpoint-NN.md — auto-generated mid-session by hooks. Incorporated and deleted by /wrapup.

Tier 3 — Semantic memory (05-agent/MEMORY.md + 05-agent/INDEX.md) Always loaded at session start. MEMORY.md holds Identity, Active Projects, and Critical Behaviors — no hard limit, ~55 lines target (3 sections only). INDEX.md lists all memory files in memory/ and is loaded alongside MEMORY.md. Only /learn writes to MEMORY.md Critical Behaviors. Use /doctor --fix to audit and repair stale entries.

Tier 4 — Knowledge base (03-knowledge/, 05-agent/memory/) Permanent, searchable notes. /distill crystallizes a completed topic thread into a structured note in 03-knowledge/. /learn saves behavioral patterns and domain facts to memory/. /recap batch-promotes from session logs to memory/ — does NOT write to MEMORY.md.

Task Syntax

OneBrain uses the Obsidian Tasks plugin format:

- [ ] Task description 📅 2026-03-25
- [ ] High priority task 🔺 📅 2026-03-22

Tasks live inline in your notes — the Tasks plugin surfaces them across the vault. Run /tasks to open a live dashboard in Obsidian (TASKS.md at vault root) with sections for overdue, due this week, unscheduled, due later, and recently completed.

Prerequisites

Required: git — used to version-control your vault.

Verify with git --version before running the installer.

Recommended for auto-checkpoints: Python or Node.js — used by the checkpoint hook to build JSON. The hook falls back to a pure-bash implementation if neither is found, but having one in your PATH is more reliable. On Windows, install Python from python.org and ensure "Add to PATH" is checked during setup.

Windows: Git for Windows (above) includes Git Bash, which provides the bash environment required to run all hooks.

Community Plugins

These three plugins are pre-configured in vault settings — install them via Settings → Community plugins → Browse, then click Trust author and enable plugins when prompted:

• Tasks — task management with due dates
• Dataview — query notes like a database
• Terminal — run your AI agent from within Obsidian

These are recommended but optional:

• Templater — advanced templates
• Calendar — visual calendar view
• Tag Wrangler — manage tags across vault
• QuickAdd — fast capture workflows
• Obsidian Git — version control for your vault

Claude Code Skills (Optional)

For Obsidian-specific Claude Code skills (markdown, bases, canvas, and more), install the Obsidian Skills plugin separately:

/plugin marketplace add kepano/obsidian-skills
/plugin install obsidian@obsidian-skills