Obsidian Second Brain
Cross-CLI skill for Obsidian. Turns your vault into a living AI-first second brain across Claude Code, Codex CLI, Gemini CLI, and OpenCode. 34 commands, vault-first research, scheduled agents, write-time AI-first validator.
Last verified:
What is Obsidian Second Brain?
Obsidian Second Brain is obsidian-second-brain is an open-source Claude Code skill that transforms any Obsidian vault into a living AI-first second brain. It works across four CLI platforms: Claude Code, Codex CLI, Gemini CLI, and OpenCode. The skill ships 34 commands across four layers (vault management, thinking tools, research toolkit, and scheduled agents) plus a background agent and four cron-scheduled agents that keep your vault alive automatically.
The tool solves the institutional amnesia problem where Claude sessions start from scratch every time and Obsidian notes just sit there unused. It connects these two powerful tools by letting Claude read, write, and reason over your Obsidian vault from outside Obsidian. When you ingest a source (URL, PDF, audio, screenshot), the vault rewrites itself—existing pages get updated, contradictions resolve automatically, and patterns trigger new synthesis pages. The nightly agent closes the day, reconciles contradictions, synthesizes cross-source patterns, heals orphan notes, and rebuilds the index while you sleep.
Key features include 34 slash commands like /obsidian-save (saves decisions/tasks/people from conversations), /obsidian-ingest (rewrites 5-15 vault pages from one source), /obsidian-challenge (your vault argues against your ideas using your own history), /obsidian-emerge (surfaces unnamed patterns from 30 days of notes), /research and /research-deep (web research with citations, vault-first gap analysis), /x-read and /x-pulse (live X/Twitter access via Grok), /youtube and /podcast (transcript extraction + summary), and Google Calendar integration. The AI-first vault rule ensures every note has a
Obsidian Second Brain pricing
Pricing model: Freemium
The tool is completely free and open-source under the MIT license. No subscription or paid plans required for core vault commands (/obsidian-save, /obsidian-daily, /obsidian-ingest, /obsidian-challenge, etc.). The research toolkit has optional API keys with pay-per-use costs: /x-read ~$0.05/call via xAI Grok, /x-pulse ~$0.13/call, /research ~$0.04/call via Perplexity, /research-deep ~$0.40-$0.80/call, /youtube ~$0.04/call, /podcast ~$0.04 Grok call plus ~$0.006/min if Whisper transcription used. Without API keys, /research and /research-deep automatically fall back to free key-less sources (Wikipedia, HackerNews, arXiv, Reddit, Lobsters, dev.to, OpenAlex, Semantic Scholar, CrossRef, DuckDuckGo). The other research commands degrade gracefully with clear setup messages when keys are missing. Gemini's /notebooklm has a free tier that covers typical usage (~$0.004/call on Flash, ~$0.06/call on Pro).
Obsidian Second Brain pros
- Open-source and free to use with MIT license
- Works across 4 CLIs: Claude Code, Codex CLI, Gemini CLI, OpenCode
- 34 slash commands covering vault management, thinking, research, and scheduling
- Vault rewrites itself—existing pages update instead of just appending
- Automatic contradiction resolution with /obsidian-reconcile
- 4 scheduled agents run while you sleep (nightly, morning, weekly, health)
- Background agent fires after every context compaction
- AI-first note format optimized for future-Claude retrieval
- Live X/Twitter access via /x-read and /x-pulse with Grok integration
- Vault-first research with /research-deep scans your notes before web search
- Bi-temporal facts track when something was true and when vault learned it
- Works offline against local filesystem with no MCP server required
- No vendor lock-in—everything stored as plain markdown files
- 4 role presets (executive, builder, creator, researcher) with tailored structures
- /obsidian-challenge pushes back using your own past failures and decisions
- /obsidian-emerge finds unnamed patterns across 30 days of notes automatically
- Ingests URLs, PDFs, audio files, and screenshots—transforms them into knowledge
- /notebooklm uses Gemini File Search for vault-grounded synthesis without browser
- Research commands work without API keys using free key-less sources
- Auto-open Obsidian after every save for immediate visibility
Obsidian Second Brain cons
- Research toolkit primarily tested on macOS, Windows/Linux support incomplete
- 3 Google Calendar commands work only on Claude Code platform
- Requires API keys for full research toolkit functionality (xAI, Perplexity, Gemini)
- No automated test suite yet—verification is manual
- No isolation within a single vault (no --scope on commands yet)
- Install.sh and auto-open behavior assume macOS conventions
- Spotify URLs not supported due to DRM restrictions
- Costs add up for heavy research usage (~$0.40-$0.80 per /research-deep call)
- Steep learning curve for understanding the 4-layer architecture
- AI-first notes designed for Claude not humans—may feel unnatural to read
- No GUI—everything happens through CLI which may deter non-technical users
- Requires git and bash for installation and updates
- Background agent adds another headless claude process that runs continuously
- Single maintainer (Eugeniu Ghelbur) with limited contributor base (9 contributors)
- Auto-synthesis may produce unwanted notes if not monitored
- No hard spend caps on API usage—user must monitor their own spend
Frequently asked questions about Obsidian Second Brain
What is a Claude Code skill?
A Claude Code skill is a reusable behavior package for Anthropic's Claude Code CLI. It bundles slash commands, scripts, references, and operating instructions that Claude loads automatically. Skills give Claude domain expertise without prompt-engineering each session. This obsidian-second-brain skill specifically gives Claude the ability to read, write, and reason over your Obsidian vault.
Is this an Obsidian plugin or a Claude Code skill?
This is a Claude Code skill, not an Obsidian plugin. An Obsidian plugin lives inside Obsidian and adds UI features there. A Claude Code skill lives inside Claude Code (Anthropic's terminal AI coding agent) and gives Claude the ability to read, write, and reason over your Obsidian vault from outside Obsidian. You install this skill into Claude Code, not into Obsidian. Your vault is unchanged, just better-leveraged.
How do I install this skill?
Run the one-line installer: curl -fsSL https://raw.githubusercontent.com/eugeniughelbur/obsidian-second-brain/main/scripts/quick-install.sh | bash. Or use the two-command method: git clone https://github.com/eugeniughelbur/obsidian-second-brain ~/.claude/skills/obsidian-second-brain, then bash ~/.claude/skills/obsidian-second-brain/scripts/setup.sh '/path/to/your/vault'. Restart Claude Code after install and run /obsidian-init to generate the vault structure.
Does this work with Codex CLI, Gemini CLI, or OpenCode?
Yes. The repo ships a build script that compiles the platform-neutral source into four platform-specific outputs. Run bash scripts/build.sh --platform codex-cli (or gemini-cli, or opencode), then copy the resulting dist/<platform>/ tree into your vault. The non-Claude builds auto-generate a routing table that maps natural-language triggers to command files, so 35 cross-platform commands work on any CLI (the 3 Google Calendar commands are Claude Code only).
Do I need API keys to use this?
Mostly no. The vault commands (/obsidian-save, /obsidian-daily, etc.) need no API keys. /research and /obsidian-deep are also key-free—with no Perplexity key they automatically fall back to free, key-less sources (Wikipedia, HackerNews, arXiv, Reddit, and more). The remaining research commands (/x-read, /x-pulse, /notebooklm, /youtube, /podcast) need their respective keys and exit with a clear setup message when one is missing. The calendar commands need the Google Calendar MCP connector rather than an API key.
What is the AI-first vault rule?
The principle that vault notes are written for future-Claude to retrieve and reason over, not for human reading. Notes have a mandatory '## For future Claude' preamble at the top, rich frontmatter (type, date, tags, ai-first: true, plus type-specific fields), [[wikilinks]] for every person/project/idea referenced, external claims carry recency markers like '(as of 2026-04, source.com)', source URLs preserved verbatim inline, and confidence levels where applicable (stated | high | medium | speculation).
What does /research-deep do that /research doesn't?
/research runs a single Perplexity query and returns a dossier with citations. /research-deep is vault-first: it scans your existing notes, identifies what you already know about the topic, spawns 3-5 targeted follow-up searches to fill only the gaps, and produces a delta report showing what's new, what's confirmed, contradictions to resolve, and recommended vault updates. Vault-first means you stop re-researching what's already in your notes.
Is this safe to run on my existing vault?
Yes. The skill never deletes or modifies notes destructively without explicit confirmation. Existing notes stay as-is. New notes follow the AI-first rule. /obsidian-health flags pre-AI-first notes so you can update them on your own schedule. The skill writes to your vault as standard markdown files, so Obsidian Sync, iCloud, Syncthing, and Git-based sync all work without modification.
How is this different from Notion AI or Mem?
Notion AI and Mem are closed-source SaaS products that own your data. This skill stores everything as plain markdown in your local Obsidian vault, with no vendor lock-in. The AI is on top of your data, not behind it. You can switch tools or stop using the skill at any point and still have your full vault. You also get scheduled agents that run while you sleep and thinking tools that challenge your ideas.
How do I update to the latest version?
Run: cd ~/.claude/skills/obsidian-second-brain && git pull. Nothing to re-run—commands pick up the new instructions automatically. See CHANGELOG.md in the repo for what's in each release. The skill is actively maintained with new commands and improvements added regularly (latest v0.8.0 rewrote /notebooklm to use Gemini File Search API).