Version: 1.0 Created: 2025-11-19 Architecture: PAI v1.2.0 + AI-Agnostic + Obsidian Integration
The Flourisha AI Brain is your centralized AI infrastructure that:
- ✅ Works with any AI (Claude, Gemini, Copilot, etc.)
- ✅ Syncs with Google Drive for multi-device access
- ✅ Editable in Obsidian for beautiful knowledge management
- ✅ Follows PAI best practices (Skills-as-Containers)
- ✅ Integrates with PARA methodology (
00_= foundation)
00_AI_Brain/
├── README.md # This file
├── skills/ # PAI Skills (symlinked to AI vendors)
│ ├── CORE/ # Flourisha identity
│ ├── research/ # Research workflows
│ ├── fabric/ # Fabric patterns
│ └── [other skills]/
│ ├── SKILL.md # Skill definition
│ ├── workflows/ # Specific workflows
│ ├── assets/ # Templates, resources
│ ├── examples/ # Example outputs
│ └── scripts/ # Helper scripts
├── docs/ # System documentation
│ ├── README.md # Documentation index
│ ├── startup/ # Startup procedures
│ ├── security/ # Security docs
│ ├── monitoring/ # Monitoring guides
│ ├── mcp-servers/ # MCP documentation
│ └── [other categories]/
├── scripts/ # System automation
│ ├── startup/ # Service startup scripts
│ ├── security/ # Security scripts
│ ├── monitoring/ # Monitoring scripts
│ └── backup/ # Backup scripts
└── context/ # AI context files
├── MASTER_CONTEXT.md # System overview
└── [other contexts]/
Read this first:
/root/flourisha/00_AI_Brain/context/MASTER_CONTEXT.mdThen explore:
- Skills:
/root/flourisha/00_AI_Brain/skills/ - Docs:
/root/flourisha/00_AI_Brain/documentation/README.md - Scripts:
/root/flourisha/00_AI_Brain/scripts/
In Terminal:
# Navigate to AI Brain
cd /root/flourisha/00_AI_Brain
# Browse skills
ls -la skills/
# Read master context
cat context/MASTER_CONTEXT.md
# View docs index
cat docs/README.mdIn Obsidian (Windows):
- Open Obsidian
- Navigate to
00_AI_Brain/ - Edit skills, docs, or context files
- Changes auto-sync via Google Drive
Every skill follows this pattern:
skill-name/
├── SKILL.md # Core definition (Tier 2)
│ ├── Frontmatter # name, description, triggers (Tier 1)
│ └── Body # Instructions, workflows
├── workflows/ # Specific task workflows (Tier 3)
│ ├── main.md
│ └── advanced.md
├── assets/ # Templates, resources (Tier 3)
│ └── template.md
├── examples/ # Example outputs (Tier 3)
│ └── example.md
└── scripts/ # Helper scripts (Tier 3)
└── helper.sh
Tier 1: Metadata (always loaded, ~100 tokens)
- Skill name, description, triggers
- Loaded at startup for routing
Tier 2: Instructions (loaded when triggered, ~2000 tokens)
- Full SKILL.md with instructions
- Available workflows
- How to use the skill
Tier 3: Resources (loaded as needed, 500-2000 tokens each)
- Individual workflow files
- Assets and templates
- Examples and scripts
# Navigate to skills
cd /root/flourisha/00_AI_Brain/skills
# Create new skill structure
mkdir -p new-skill/{workflows,assets,examples,scripts}
# Create SKILL.md
cat > new-skill/SKILL.md << 'EOF'
---
name: new-skill
description: |
Brief description.
USE WHEN user says 'trigger phrase', 'keyword'
---
# New Skill
## Available Workflows
- workflows/main.md - Primary workflow
## Assets
- assets/ - Supporting resources
## Examples
- examples/ - Example outputs
EOF
# Create example README
echo "# Examples for new-skill" > new-skill/examples/README.mdSkills live here once:
/root/flourisha/00_AI_Brain/skills/
AI vendors access via symlinks:
# Claude
/root/.claude/skills → /root/flourisha/00_AI_Brain/skills/
# Gemini (future)
/root/.gemini/skills → /root/flourisha/00_AI_Brain/skills/Benefits:
- ✅ Update skill once, all AIs see it
- ✅ Edit in Obsidian, syncs to server
- ✅ Automatic Google Drive backup
- ✅ Works with any AI tool
Each AI has its own config directory:
- Claude:
/root/.claude/(hooks, agents, commands) - Gemini:
/root/.gemini/(future configs)
Shared resources (skills, docs, scripts) in AI Brain.
Locations:
- Server:
/root/flourisha/00_AI_Brain/ - Google Drive:
Flourisha_gDrive/00_AI_Brain/ - Obsidian (Windows):
G:\Shared drives\Flourisha_gDrive\00_AI_Brain\
Sync Commands:
# Pull from Google Drive
flourisha-pull
# Push to Google Drive
flourisha-push
# Check Drive sync status
rclone lsd Flourisha_gDrive: | grep 00_AI_BrainWorkflow:
- Open Obsidian on Windows
- Navigate to
00_AI_Brain/ - Edit any skill, doc, or context file
- Save (auto-syncs to Google Drive)
- Google Drive syncs to server
- AI sees updated content immediately
Benefits:
- Beautiful markdown editor
- Graph view of relationships
- Powerful search across all docs
- Templates for new skills
- Multi-device access
documentation/DOCUMENTATION_MAP.md - Master documentation index
| Service | Description |
|---|---|
| Document Processor | Extract text from PDFs, images |
| Knowledge Ingestion | Pipeline to all stores |
| Extraction Backends | Claude vs Docling backends |
| Store | Description |
|---|---|
| Three-Store Overview | Vector + Graph + Whole |
| Database Schema | Supabase tables |
| Vector Store | pgvector embeddings |
| Graph Store | Neo4j/Graphiti |
- Startup:
documentation/startup/services.md - Security:
documentation/security/scanning.md - Monitoring:
documentation/monitoring/overview.md
context/MASTER_CONTEXT.md - System overview
- Quick facts
- Key services
- Navigation
- Contact info
scripts/startup/start_services_lean.py
- Selective service startup
- Full documentation in file
Usage:
# Start everything
python3 /root/flourisha/00_AI_Brain/scripts/startup/start_services_lean.py
# Start specific service
python3 /root/flourisha/00_AI_Brain/scripts/startup/start_services_lean.py --services neo4j
# List available services
python3 /root/flourisha/00_AI_Brain/scripts/startup/start_services_lean.py --list- Security:
scripts/security/- Security scanning, audits - Monitoring:
scripts/monitoring/- Health checks, monitoring - Backup:
scripts/backup/- Backup automation
On Server:
cd /root/flourisha/00_AI_Brain/skills/skill-name
nano SKILL.md
flourisha-push # Sync to Google DriveIn Obsidian:
- Edit skill in Obsidian
- Save (auto-syncs)
- Done!
On Server:
cd /root/flourisha/00_AI_Brain/documentation/category
nano new-doc.md
flourisha-pushIn Obsidian:
- Create new note in docs/
- Write documentation
- Auto-syncs to server
Google Drive provides automatic versioning:
- File history available in Google Drive
- Can restore previous versions
- No manual git commits needed for docs/skills
- Skills stored once in AI Brain
- Symlinked to vendor configs
- Edit once, works everywhere
- Small metadata always loaded
- Full instructions loaded on trigger
- Resources loaded as needed
- Shared resources in AI Brain
- Vendor-specific configs separate
- Clean boundaries
00_AI_Brain= System foundation- Part of Flourisha knowledge system
- Syncs with projects, areas, resources
# Navigate
cd /root/flourisha/00_AI_Brain
# List skills
ls -la skills/
# Read master context
cat context/MASTER_CONTEXT.md
# View docs
cat docs/README.md
# Sync to Google Drive
flourisha-push
# Pull from Google Drive
flourisha-pull
# Search everything
grep -r "search term" /root/flourisha/00_AI_Brain/- Always read
context/MASTER_CONTEXT.mdfirst - Reference specific docs as needed
- Don't load everything at once
- Browse in Obsidian for best experience
- Use terminal for quick edits
- Google Drive keeps everything backed up
- v1.0 (2025-11-19) - Initial AI Brain structure
- Skills-as-Containers (PAI v1.2.0)
- AI-agnostic architecture
- Google Drive + Obsidian integration
- Multi-vendor support via symlinks
Questions? All documentation is in documentation/