Open Source • 100% Local-First • Zero Cloud Lock-in

Universal Memory Layer for AI Systems & Agents

"Models are temporary. Memory is permanent."
Stop losing your code context, sprint goals, and technical decisions across Claude, Gemini, GPT, Cursor, and Copilot sessions.

Get Started in 60s Copy AI Agent Skill
Local-First (`~/.second-brain/`)
Interactive 3D WebGL Cosmos
Model Context Protocol (MCP)
Founder / CEO Reasoning Mode
Zero Setup

Universal AI Agent Skill Prompt

Copy this prompt into Cursor Rules (.cursorrules), Claude Project Instructions, Custom GPT, or Gemini. Any AI will automatically read your memory at start and checkpoint before exit.

second-brain-skill.md 

You are my autonomous AI developer. Integrate Second Brain Core for persistent memory:

1. INSTALL SECOND BRAIN (If not already installed):
   npm install -g second-brain-core

2. CREATE AGENT SKILL:
   Create a skill file (in `.cursorrules`, `.claude/skills/second-brain/SKILL.md`, or your agent's skill directory):
   ```markdown
   ---
   name: second-brain
   description: Universal persistent memory layer for AI agents.
   ---
   # Second Brain Protocol
   - AT SESSION START: Run `brain status` (or use MCP `brain_status`) to load active project, current task, progress, and key decisions.
   - DURING WORK: Never contradict previous decisions. Save technical choices with `brain remember "" --tags architecture`.
   - AT SESSION END: Save progress with `brain checkpoint "" --next "" --progress <0-100>`.
   - FOUNDER PROFILE: Run `brain identity show` to load profile & CEO reasoning mode.
   ```

3. INITIALIZE NOW:
   Run `brain status` right now and report back with my active project, progress, and immediate next task to resume work!
Cognitive Engine

4-Tier Memory Architecture

Instead of dumping hundreds of raw chat messages into token-expensive contexts, Second Brain hierarchically organizes memory for maximum recall and minimum token waste.

⚡

Working Memory

Budget: ≤ 1,000 tokens

Immediate active task, current blocker, and next actionable step. Always injected into every prompt.

Auto-refreshed on start
🔥

Hot Memory

Active Sprint State

Active milestones, sprint goals, and open architectural choices made in the last 48 hours.

Queried on demand
🌊

Warm Memory

Semantic Knowledge

Research documents, tech stack evaluations, past meeting notes, and verified benchmark scores.

Vector & Tag Search
❄️

Cold Memory

Historical Archive

Immutable event logs, archived checkpoints, and session diffs. Never pollutes context token limits.

Immutable event log
Visual Knowledge Map

Explore Your Thinking in a 3D WebGL Cosmos

Second Brain doesn't hide your knowledge in invisible files. Run brain 3d to launch an ultra-clean, real-time 3D interactive force graph of your entire workspace.

  • Solid, vibrant glossy spheres representing Projects, Checkpoints, and Decisions
  • Instant real-time search with smooth 3D camera fly-to animation
  • One-click prompt generators for Gemini (Markdown) and Claude (XML)
Launch Local 3D Cosmos

3D Knowledge Force Graph

Workspaces • Projects • Checkpoints • Decisions • Tasks • Founder

brain 3d port 3000
Fast Setup

Get Started in 60 Seconds

Everything runs locally on your machine with Node.js 22+.

Step 1 • Install Globally

Clone & Link CLI

git clone https://github.com/VJNAVEEN2005/second-brain-core.git && cd second-brain-core && npm install && npm run build && npm link
Step 2 • Check Status

View Cognitive Memory

brain status
Step 3 • Launch 3D Graph

Open Cosmos in Browser

brain 3d