obsidian-second-brain
Persistent memory for Claude Code and 6 other CLI agents, stored as plain markdown in your Obsidian vault. Stop re-explaining your projects, decisions and people every session. 45 commands: hybrid semantic search, self-rewriting notes, key-less web research, and scheduled agents that maintain the va
obsidian-second-brain is a Claude Code plugin and cross-platform agent skill that provides persistent memory for AI coding assistants by storing context as plain markdown in your Obsidian vault. It enables Claude Code, Codex, Gemini CLI, OpenCode, Grok Bot, Hermes, Pi, and other AI agents to remember project decisions, people, and conversations across sessions, eliminating the need to re-explain context every time. With 47 commands, it offers hybrid semantic search, self-rewriting notes, API-key-free web research, and scheduled agents that automatically maintain your vault, turning your Obsidian notes into an AI-accessible second brain.
Key Features
Use Cases
- 01Developers maintaining project memory: automatically save architecture decisions, code discussions, and API design choices to vault notes that persist across Claude Code sessions
- 02Founders tracking decisions and people: capture meeting outcomes, relationship context, and strategic pivots in daily notes that AI agents can recall months later
- 03Researchers building knowledge bases: ingest PDFs, YouTube videos, and web articles that automatically rewrite existing concept notes and resolve contradictions
- 04Content creators planning topics: use /x-pulse to scan trending X themes, /research-deep to fill vault knowledge gaps, and /obsidian-challenge to test ideas against past work
- 05Teams with multi-repo workflows: configure per-project vault paths so each repository's Claude Code session uses its own dedicated second brain
- 06Privacy-focused users: run entirely offline with local Ollama embeddings, Whisper transcription, and key-free research fallbacks (Wikipedia, arXiv, HackerNews)
Related Plugins
View moreandrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
claude-mem
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Understand-Anything
Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.
rtk
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
obsidian-second-brain — FAQ
What is obsidian-second-brain and how does it relate to Obsidian plugins?+
This is a Claude Code plugin (not an Obsidian plugin) that gives AI coding agents like Claude Code, Codex, and Gemini CLI the ability to read, write, and reason over your Obsidian vault from the command line. It stores persistent memory as plain markdown files in your existing vault, enabling AI sessions to remember context across conversations.
How do I install the Claude Code plugin version?+
Inside any Claude Code session, run '/plugin marketplace add eugeniughelbur/obsidian-second-brain' then '/plugin install obsidian-second-brain@obsidian-second-brain'. Add your vault path to ~/.claude/settings.json under the env section (OBSIDIAN_VAULT_PATH), restart Claude Code, and run '/obsidian-second-brain:obsidian-init' to set up the vault structure.
Does this work with other AI coding assistants besides Claude Code?+
Yes, it supports Codex CLI, Gemini CLI, OpenCode, Grok Bot, Hermes, Pi, Google Antigravity, and any Agent Skills-compatible harness. Run 'bash scripts/build.sh --platform <name>' to generate the platform-specific build, then copy the dist/<platform>/ files to your vault. The 44 cross-platform commands (47 total, but /obsidian-calendar is Claude Code-only) work identically across all platforms.
Do I need API keys to use this?+
Most commands need no API keys—vault operations (/obsidian-save, /obsidian-daily, /obsidian-find, etc.) work out of the box. The research toolkit commands (/x-read, /x-pulse, /youtube, /notebooklm, /podcast) require API keys for xAI Grok, Perplexity, or Gemini, but /research and /research-deep now fall back to free key-less sources (Wikipedia, arXiv, HackerNews, Reddit) when no Perplexity key is set.
Is this free and open source?+
Yes, the plugin is MIT-licensed and free to use. The core vault commands cost nothing; optional research toolkit commands use pay-per-use APIs (xAI Grok ~$0.05/call, Perplexity ~$0.02-$0.50/call, Gemini free tier or ~$0.004-$0.06/call). Semantic search can run entirely locally via free Ollama models (bge-m3) with no external API calls.
How does this keep my vault up to date without me doing anything?+
Scheduled agents run automatically: a morning agent creates daily notes with calendar events and overdue tasks, a nightly agent closes the day and reconciles contradictions, a weekly agent runs structured reviews, and a health agent audits the vault. An optional background agent (opt-in, runs after context compaction) updates notes while you work. All agents follow the OKM freshness rule: facts are timeless, dated, or pointers to live sources.
How do I install obsidian-second-brain?+
Open the source repository on GitHub and follow its README. obsidian-second-brain is a plugin — MCP Agents Market links you directly to the official repo.
Is obsidian-second-brain free?+
obsidian-second-brain is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.