Second Brain Frameworks
Proven frameworks for building an AI-powered second brain — from the foundational CODE workflow to the seven subsystems that separate a living brain from a digital graveyard.
← Back to What Is a Second Brain?The CODE Framework
Capture → Organize → Distill → Express
Originally developed by Tiago Forte, CODE is the foundational workflow for any second brain. AI transforms every step — especially Distill, which was historically the bottleneck that killed most PKM systems.
Capture
Everything, frictionlessly. Voice memos, web clips, PDFs, emails. "Capture everything, organize nothing" initially — friction kills the habit.
Organize
By utility, not origin. AI routes content into the PARA structure automatically — no manual tagging required. The system structures itself.
Distill
AI synthesizes 100 notes, a 60-minute transcript, or a 30-page document into actionable insight in seconds. This is where most manual systems fail — and where AI wins.
Express
Never start blank. Assemble pre-distilled building blocks. Write faster, decide better, create from a foundation instead of from scratch.
The PARA Method
Projects · Areas · Resources · Archives
PARA organizes everything you capture by time horizon and actionability — not by subject or source. It is the organizational backbone of the CODE framework, now automated by AI tagging.
Projects
Deadline-driven outcomes. Active work with a defined end state. These get your attention now. Days to weeks.
Areas
Ongoing responsibilities with no end date — health, finances, team management, professional development. Always alive.
Resources
Topics of ongoing interest — reference material, research, reading lists. You access these when relevant, not constantly.
Archives
Completed projects and inactive items. AI search makes archiving safe — nothing is truly lost, just out of the way.
Three Operating Rooms
Library → Lab → Factory
Every well-built second brain has three zones. Most people only build the first one — the filing cabinet. The Factory is where knowledge stops being passive and starts driving action.
The Library
Capture & Storage. AI-native ingestion from voice memos, web clips, PDFs, emails, and meeting transcripts — without manual tagging. Where information enters the system.
The Lab
Insights & Discovery. Knowledge graphs, semantic search, and pattern recognition surface relationships between ideas you could never find manually. Where connections are made.
The Factory
Automation & Action. Where knowing becomes doing. Agents live inside your knowledge base and execute tasks — triggered by context, not by you.
The 7 Core Subsystems
The minimum viable architecture for a living second brain
Both Cole Medin's conversational second brain and JARVIS converged on the same seven subsystems — strong evidence this is the right architecture pattern regardless of implementation style or language.
Memory Layer
Store, index, and retrieve everything. SQLite + full-text search as the baseline; vector embeddings for semantic similarity.
Ingestion Pipeline
Bring new information into the system from wherever you already work — APIs, web crawlers, file uploads, voice transcription.
Proactive Intelligence
Monitor and alert autonomously via scheduled heartbeat jobs. The system runs whether you ask or not.
Reflection Agent
Curate and distill over time. Daily/weekly synthesis promotes key insights to long-term memory and prunes noise.
Chat Interface
Conversational access to your knowledge — Slack bot, local web UI, or CLI. How you ask the brain questions.
Skills System
Extend capabilities via markdown-based SKILL files. Add new behaviors without touching the core codebase.
Scheduler
Prevent knowledge decay. Weekly re-index, link checking, staleness detection. What separates a living brain from a digital graveyard.
See These Frameworks in Practice
JARVIS implements all of these frameworks. Watch it being built from scratch on YouTube, or explore the case study to see how they fit together in a real system.