Second Brain: Claude Code + Obsidian

second-brain

Second Brain: Claude Code + Obsidian

Wire Claude Code into your second brain. Obsidian, PARA, daily notes, project archives: all editable, all queryable.

The Problem

Your Notes Are Passive. They Don't Have to Be.

Most people treat Obsidian as a filing cabinet. Notes go in, rarely come out, never connect to the work you're doing in Claude Code. The two systems sit side by side, completely separate, and the intelligence is lost in the gap.

What passive notes cost you

• Re-explaining client context to Claude every single session
• Knowledge trapped in Obsidian that Claude can never access
• Manual copy-paste between your vault and your AI workflows
• Meeting notes that never get processed into tasks or CRM updates

PARA

Folder Structure

Auto-Route

Inbox Processing

Persistent

Memory System

Live Sync

Obsidian Graph

Why Integrate

Claude + Obsidian Is a Force Multiplier

When Claude Code runs inside your vault, every .md file becomes both human-readable and machine-queryable. You get a second brain that actually thinks back.

Bidirectional Read/Write

Claude reads your notes to get context, then writes results back as new notes. Meeting recap lands in the client folder automatically. Graph view updates instantly.

Persistent Client Context

Every client gets a CLAUDE.md in their folder. When you work on that client, Claude reads the scoped context. No re-explaining stack, engineers, or conventions.

[[Wikilinks]] at Scale

Claude writes [[wikilinks]] between related notes. Obsidian renders them instantly in graph view. A knowledge map builds itself without any manual effort.

Compounding Memory

Session 1 you explain everything. Session 50 Claude already knows your vault structure, client names, file conventions, and preferences. Zero ramp-up time.

Architecture

How Claude Lives Inside Your Vault

Your Obsidian vault is Claude's working directory. The filesystem IS the integration: no plugins, no APIs, no sync layer.

Obsidian Vault = Claude's Working Directory

CLAUDE.md -- full identity, PARA structure, preferences

MEMORY.md -- Claude's own notes across sessions

00_Inbox/ 01_Projects/ 02_Areas/ 03_Resources/

.claude/skills/ -- read/write vault files

.claude/agents/ -- subagents for tasks

*.md files -- shared between Claude + Obsidian

[[wikilinks]] -- appear in Obsidian graph view

External Integrations: All Flow Back as .md Files

PARA Method + Claude

Every Folder Has a Purpose. Claude Knows Each One.

PARA (Projects, Areas, Resources, Archive) gives your vault a clear grammar. Claude learns the grammar and routes everything correctly without you deciding where each file goes.

Raw notes, ideas, meeting dumps, screenshots, voice memos transcribed. Everything lands here first.

What Claude Does

Reads every file daily. Extracts action items. Routes to the correct PARA folder. Creates CRM tasks. Archives processed items. Nothing stays in Inbox more than 24 hours.

Client projects, internal builds, product launches. Each has a clear start, end, and deliverables. Structured as Clients/Active, Clients/Prospects, Internal/, Products/.

What Claude Does

Updates Progress.md with sprint logs. Appends to Communication-Log.md after meetings. Tracks blockers and unresolved items. Links related resources from 03_Resources/. Generates weekly status summaries per client.

Sales, Marketing, Operations, Technical, Legal. These never end. They are areas of continuous responsibility. SOPs, metrics, strategies, team docs.

What Claude Does

Maintains SOPs when processes change. Updates ICP and sales strategy docs. Generates content drafts in 02_Areas/Marketing/. Tracks operational metrics. Keeps Legal/ compliance docs current.

AI agent architectures, n8n workflows, automation patterns, tool databases, learning notes, prompt libraries. Knowledge you reference repeatedly.

What Claude Does

Builds INDEX.md files with [[wikilinks]] to connect related concepts. Finds conceptual links across files. Creates Maps of Content (MOCs). Adds TL;DR sections to research notes.

Finished client engagements, deprecated SOPs, old experiments. Searchable but out of active view.

What Claude Does

Moves completed projects here automatically. Preserves all history. Searches archive when similar problems surface in active projects. Extracts patterns from past work to inform current builds.

Setup

Vault Setup: 5 Steps

No special plugins. Five configuration steps turn your existing Obsidian vault into an active intelligence layer.

Create a CLAUDE.md at the root of your vault. This tells Claude who you are, what your folder structure looks like, and how to behave.

Add rules that apply to specific parts of your vault. These are granular instructions Claude follows when working in each PARA area.

Create a skill that processes your inbox. This is the highest-leverage first automation. It keeps your vault clean daily with one command.

Auto-memory is enabled by default. Review the settings to understand what gets stored and where. Claude saves patterns across sessions.

Open Claude Code in your vault directory and run the first command. Watch it read your CLAUDE.md, understand your structure, and start working.

Daily Notes

Daily Briefing: Auto-Generated Every Morning

Claude combines your calendar, CRM state, vault updates, and recent client activity to generate a morning briefing. Priorities, meetings, blockers: all in one file.

Read today's calendar events via gogcli

Check Attio CRM for overdue tasks and upcoming deadlines

Scan recent changes across 01_Projects/Clients/Active/

Identify unresolved blockers from Communication-Log.md files

Generate prioritized daily brief and write it to 00_Inbox/daily-brief-[date].md

Query Patterns

5 Key Workflows for an Active Vault

Each workflow turns passive files into working systems. These are the commands you'll run daily once the vault is wired.

Every morning: Claude reads 00_Inbox/, extracts action items from each note, routes to the correct PARA folder, creates tasks in Attio, archives processed items.

Drop a meeting transcript into your vault. Claude reads it, extracts decisions and actions, updates the client's Communication-Log.md, and creates CRM tasks.

Claude reads all files in a resource folder, finds conceptual connections, and creates an INDEX.md with [[wikilinks]]. Links appear instantly in Obsidian graph view.

Claude reads research notes, distills the most valuable insights, adds TL;DR sections, links to source notes, and reduces noise. Your vault gets smarter without getting bigger.

Generate a current-state snapshot for any client. Pulls from README, Progress, Communication-Log, and Technical docs. One command for a full client briefing.

Worked Use Cases

3 Real Scenarios, Step by Step

These are the three highest-impact integrations to run first. Each one pays for the setup within the first week.

Use Case 01

Auto-Routing an Inbox Full of Meeting Notes

Scenario

You have 12 unprocessed meeting notes from the last week across 4 different clients.

Run claude /inbox-processor from the vault root

Claude reads all 12 files, identifies each client by matching keywords against 01_Projects/Clients/Active/

Each note is appended to the correct Communication-Log.md

Action items extracted and pushed to Attio as tasks

Inbox is cleared and a PROCESSED.md summary is written

Result

12 files routed, 18 Attio tasks created, 0 manual decisions required.

Use Case 02

Building a Knowledge Map for AI Agent Research

Scenario

You have 30+ notes on AI agent architectures scattered across Resources/. No index, no connections.

Run: claude 'Read all files in 03_Resources/AI_Agents/ and build a knowledge map'

Claude scans all 30 files and extracts key concepts

Finds overlapping concepts: tool-calling, memory, orchestration, MCP

Creates INDEX.md with [[wikilinks]] grouped by theme

Obsidian graph view immediately shows the connections

Result

One INDEX.md file makes 30 notes instantly navigable in Obsidian.

Use Case 03

Per-Client CLAUDE.md for Zero-Context Sessions

Scenario

You switch between 5 clients daily. Explaining context to Claude each session wastes 15 minutes.

Create /vault/01_Projects/Clients/Active/[Client]/CLAUDE.md per client

Include: stack, engineers, conventions, active sprint, blockers

When working on a client file, Claude reads all three CLAUDE.md levels

Global prefs (no emojis) + vault structure + client specifics, all merged

Zero re-explanation. Claude already knows everything

Result

15 minutes saved per client switch. Context is always current because the file is the source of truth.

Auto-Memory System

Claude Builds Its Own Memory About Your Vault

Every session, Claude learns your patterns, preferences, and quirks. This knowledge persists across conversations in a single readable file.

Where it lives

~/.claude/projects/<repo>/memory/MEMORY.md: one file per project, scoped to the vault. Fully readable and editable.

What gets saved

Build quirks, tool preferences, client naming conventions, folder patterns, things that failed before. Example: "always use bun not npm" or "GitButler manages commits; never use git commit directly."

How it learns

Claude detects repeated corrections and patterns. If you say "no emojis" three times, it saves to MEMORY.md automatically. Next session, it already knows.

How to browse

Use /memory to list all memory files. Read them directly in Obsidian or any text editor. Edit or delete entries that are outdated.

Why it matters

Session 1: you explain everything. Session 50: Claude already knows your vault structure, client names, file conventions, and how you like things done. Zero ramp-up time.

Want this running in your stack?

ManasFlow builds AI automation systems for production teams.

Remote Control: Drive Claude From Anywhere

Second Brain Structure: PARA Done Right