🧠 Awesome Second Brain

The AI-assisted second brain that connects your notes, ideas, and tasks into persistent context for agentic workflows.

[!IMPORTANT] Awesome Second Brain is currently at v2.0.0. We’re moving fast, so expect frequent updates and some breaking changes as the project evolves. If you’re into AI-assisted productivity, we’d love your feedback.

🌪️ The Problem

Context evaporates instead of compounding. Every conversation ends, and the decisions, constraints, and rationale that went into it vanish with the tab. You keep re-making the same calls, re-explaining the same project, re-training the same model on the same facts. The work you already did is invisible to the tool meant to accelerate it.

💡 The Solution

Give your agent an Awesome Second Brain.

Awesome Second Brain (ASB) is an open-source framework that gives your AI a spotless context layer, organized knowledge, and automated processing pipelines — so your AI operates with the same goals, history, and strategic intent as you do.

Before ASB vs After ASB

📊 Features at a Glance

ASB v2.0 — Core features are stable; workflows and documentation are maturing.

✨ New in v2.0

🔄 Arcontexta Inspired Hook System

A Claude Code plugin that generates complete knowledge systems from conversation.

🔍 QMD — On-Device Vault Search

A powerful local search engine for everything you need to remember.

⚡ See It In Action

Morning kickoff:

/open-day
# → Loads North Star, active projects, open tasks, recent git changes.
# → "You have 2 active projects. The auth refactor is blocked on API contract.
#    Your 1:1 with Sarah is at 2pm — last time she flagged observability."

Capture any raw input:

/brain-dump "I've been thinking about using Redis for caching. The API is rate limited to 1000 calls/hour."
# → Extracts atomic observations: [idea] Use Redis, [fact] API limit.
# → Suggests tags: #infrastructure #caching #api
# → Stages in inbox/raw/ for later processing.

Process raw captures into structured notes:

/process
# → Scans inbox/raw/ for unprocessed notes.
# → Adds full frontmatter: domain, type, tags, description (~150 chars).
# → Flags multi-topic notes for splitting, offers `status: thinking` for scratchpads.
# → Moves ready notes to inbox/ready/ awaiting distribution.

Distribute processed notes to their permanent home:

/distribute
# → "3 notes ready. The Redis caching note → domains/Work/02_PAGES/redis-caching.md?"
# → Detects duplicates ("Absorb into existing api-rate-limits page?") and multi-topic notes (split).
# → Adds wikilinks to related pages, updates domain INDEX.md.
# → Source note kept for provenance (status: processed).

🚀 Quick Start

Automatic Install (recommended)

Setup time: ~10 minutes.

Mac / Linux:

curl -fsSL https://raw.githubusercontent.com/superuser-pal/awesome-second-brain/main/install.sh | bash

Windows (PowerShell):

irm https://raw.githubusercontent.com/superuser-pal/awesome-second-brain/main/install.ps1 | iex

The installer checks prerequisites, asks where you want your vault, installs dependencies, and configures hooks automatically.

To update: Simply run the installer command again or git pull in your vault directory.

Post-Install Setup

1. Open the vault folder in Obsidian (File → Open Vault → Open Folder as Vault).
2. Enable the Obsidian CLI in Settings → General (requires Obsidian 1.12+).
3. Verify the installation in a new terminal:
   hljs language-bash

   Copy

   obsidian --version
   

4. Launch your Agent in the vault directory:
   hljs language-bash

   Copy

   cd ~/second-brain && claude
   

5. Initialize your brain:
   hljs language-arduino

   Copy

   /setup-context
   

   Choose Level 2 for a guided setup of your goals, your first domain, and your dashboard.

New to Terminal? See SETUP.md for a step-by-step guide with troubleshooting.

Manual Install

git clone https://github.com/superuser-pal/awesome-second-brain
cd awesome-second-brain
npm run setup   # installs hook dependencies (Bun)

Optional: QMD Semantic Search

For semantic search across the vault. The npm package is @tobilu/qmd (published from tobi/qmd):

npm install -g @tobilu/qmd
qmd collection add . --name vault --mask "**/*.md"
qmd context add qmd://vault "Engineer's work vault: projects, decisions, incidents, people, reviews, architecture"
qmd update && qmd embed

📋 Prerequisites

Minimum setup (Starter tier)

Full setup (Standard tier — recommended)

Power tier

📂 Vault Structure

Awesome Second Brain's workspace is organized to keep focus sharp and context contained:

• dashboards/: Navigation hub, HOME.md, and TASKS.md (Master task list).
• domains/: Your high-level workspaces (e.g., Work, Personal). Each domain is a siloed environment.
• work/: Cross-domain shared operations — 1:1s, incidents, and reviews.
• inbox/: The entry point. Drop raw thoughts or quick tasks here to be processed.
• brain/: The "Operating System" — NORTH_STAR.md, MEMORIES.md, RULES.md, and LOGIC.md.
• plan/: Your daily and weekly rhythm files.
• bases/: Dynamic aggregators and template definitions (e.g., Prompts.base, Strategies.base).
• thinking/: Internal logs, session transcripts, and agent scratchpads.
• .claude/: The configuration layer containing system logic, commands, and subagents.

🛠️ How It Works

The Knowledge Pipeline

Capture raw information and transform it into structured knowledge automatically.

Lifecycle Hooks

Five automated hooks handle setting up your session context:

⌨️ Command Palette

Run these directly in your AI agent. Symbols in parentheses are "nicknames" for the authoritative namespaced commands.

🧩 Skills

14 domain skills bundled into ASB, loaded on demand by your agent via the Skill tool. Skills are reusable capabilities — they encapsulate references, workflows, and examples for a specific domain so your agent doesn't re-derive them every session.

🧠 Reasoning Strategies

22 pluggable reasoning strategies that turn ASB into the best brain sparring partner live in docs/strategies/, each a dormant page you can activate for a session. CoT is pre-promoted at setup. Pick one when the task calls for a specific thinking shape, or let /thinking:eval choose for you.

Standing controls (always available):

📚 Prompt Library

257 atomic prompts live dormant in docs/prompts/. Each is a self-contained reusable prompt — activate the ones you use, leave the rest shelved to avoid bloating / autocomplete. Browse with bases/Prompts.base or the prompts skill.

Management commands (from the prompts skill)

Sample use cases

Representative categories from the library — promote the ones that match your workflow:

• Analysis: analyze_claims, analyze_incident, analyze_risk, analyze_paper, analyze_sales_call, analyze_product_feedback, analyze_logs, analyze_threat_report
• Creation: create_design_document, create_coding_feature, create_keynote, create_formal_email, create_hormozi_offer, create_5_sentence_summary, create_mermaid_visualization, create_git_diff_commit
• Extraction: extract_wisdom, extract_insights, extract_recommendations, extract_main_idea, extract_references
• Improvement: improve_writing, improve_prompt, improve_report_finding
• Summarization: summarize, summarize_paper, summarize_meeting, summarize_debate
• Review: review_design, analyze_peer_work, rate_content

🎯 Why Awesome Second Brain?

1. Stop starting from zero. Session hooks load identity, projects, and decisions before you type.
2. Build once, reuse forever. Skills, commands, and frontmatter schemas turn your patterns into infrastructure.
3. Your vault is yours. Local markdown, no cloud dependency, no vendor lock-in, no black-box state.
4. Token-efficient by design. Tiered loading keeps cost low and recall high.
5. Obsidian does the UI. Your agent does the work. You don't need to live in an IDE.

📝 License

Released under the MIT License.

Version: 2.0.0 | Last Updated: 2026-04-17