Personal AI Infrastructure (PAI)

Open-source personal AI infrastructure for orchestrating your life and work

Quick Start • Documentation • Examples • Community

---

"This project exists so that the best AI in the world is not only used by a few, but can be used by everyone."

Right now the most powerful AI setups are being built inside companies with massive engineering teams. That's backwards. These tools should be available to individuals—to you—so you can build your own AI system that actually knows how you work.

That's what PAI is. It's the scaffolding for building a personal AI that gets better over time because it's yours. Not some generic chatbot. Your system, your skills, your workflows.

---

🎯 What is PAI?

PAI (Personal AI Infrastructure) is an open-source template for building your own AI-powered operating system using Claude Code.

Core Primitives:

• Skills - Self-contained AI capabilities with routing, workflows, and documentation
• Agents - Specialized AI personalities for different tasks (engineer, researcher, designer, etc.)
• Hooks - Event-driven automation that captures work, provides voice feedback, and manages state

Approach: Start clean, small, and simple. Build the scaffolding that makes AI reliable.

---

🆕 What's New in v0.9.0

Big updates! PAI is now fully platform-agnostic - your AI identity, your system.

• Observability Dashboard - Real-time agent monitoring with live charts
• Genericized Identity - Configure your DA name, it flows everywhere
• Better Configuration - Clear docs for all environment variables

👉 See full changelog below

---

🔄 PAI vs Kai: What You Get

PAI (this repository) provides:

• ✅ Skills/agents/hooks architecture
• ✅ CORE documentation and routing
• ✅ History system (UOCS) for automatic documentation
• ✅ Example skills (research, fabric, etc.)
• ✅ Voice server skeleton
• ⚙️ Requires: API key configuration per skill

Kai (Daniel's private system) adds:

• 🔒 Personal data, contacts, and history
• 🔒 Additional private skills and workflows
• 🔒 Customized agent personalities and voices
• 🔒 Production integrations and automations

Think of it this way: PAI is the scaffolding. You build your own "Kai" on top of it.

After setup, PAI should:

• ✅ Execute hooks without errors
• ✅ Load CORE context at session start
• ✅ Route skills correctly
• ✅ Capture session history
• ✅ Launch agents successfully

Not working? Run the health check:

bun ${PAI_DIR}/hooks/self-test.ts

See PAI_CONTRACT.md for complete details on what's guaranteed vs what needs configuration.

---

🚀 Quick Start

1. Install Prerequisites

# Install Bun (PAI's package manager)
curl -fsSL https://bun.sh/install | bash

# Install Claude Code
# Follow instructions at: https://code.claude.com

2. Clone and Configure

# Clone the repository
git clone https://github.com/danielmiessler/Personal_AI_Infrastructure.git
cd Personal_AI_Infrastructure

# Copy environment template
cp .claude/.env.example .claude/.env

# Edit .env and add your API keys
# At minimum: ANTHROPIC_API_KEY=your_key_here

3. Copy to Your System

# Backup existing Claude config (if any)
[ -d ~/.claude ] && mv ~/.claude ~/.claude.backup.$(date +%Y%m%d_%H%M%S)

# Copy PAI's .claude directory to your home
cp -r .claude ~/.claude

Alternative: Symlink (keeps PAI updatable via git pull)

[ -d ~/.claude ] && mv ~/.claude ~/.claude.backup.$(date +%Y%m%d_%H%M%S)
ln -s $(pwd)/.claude ~/.claude

4. Start Claude Code

# PAI loads automatically via the CORE skill
claude

That's it! The CORE skill loads at session start and provides all PAI functionality.

📚 For detailed setup: See docs/QUICKSTART.md

---

📚 Documentation

All documentation lives in the CORE skill (.claude/skills/CORE/):

Essential Reading

• CONSTITUTION.md - System philosophy, architecture, and operating principles
• SKILL.md - Main PAI skill with identity, preferences, and quick reference
• SKILL-STRUCTURE-AND-ROUTING.md - How to create your own skills

System Guides

• hook-system.md - Event-driven automation
• history-system.md - Automatic work documentation
• VOICE.md → voice-server/README.md - Text-to-speech feedback

Reference

• prosody-guide.md - Voice emotion system
• prompting.md - Prompt engineering patterns
• terminal-tabs.md - Terminal management

---

🎨 Examples

Explore example skills in .claude/skills/:

• observability/ - Real-time agent monitoring dashboard with WebSocket streaming, live pulse charts, and event timelines
• brightdata/ - Four-tier progressive web scraping with automatic fallback (WebFetch → cURL → Playwright → Bright Data MCP)
• fabric/ - Integration with Fabric pattern system (242+ AI patterns)
• research/ - Multi-source research workflows
• create-skill/ - Templates for creating new skills
• alex-hormozi-pitch/ - Business pitch generation
• ffuf/ - Web fuzzing and security testing

Each skill demonstrates the skills-as-containers pattern with routing, workflows, and self-contained documentation.

---

🏗️ Architecture

PAI is built on three foundational principles:

1. Command Line First

Build deterministic CLI tools, then wrap them with AI orchestration. Code is cheaper, faster, and more reliable than prompts.

2. Skills as Containers

Package domain expertise in self-activating, self-contained modules. Natural language triggers automatic routing to the right skill.

3. Progressive Disclosure

Load context only when needed (3 tiers):

• Tier 1: System prompt (always active, 200-500 words)
• Tier 2: SKILL.md (on-demand, comprehensive reference)
• Tier 3: Reference files (just-in-time, deep dives)

Complete architecture: See .claude/skills/CORE/CONSTITUTION.md

---

🛠️ Technology Stack

• Runtime: Bun (NOT Node.js)
• Language: TypeScript (NOT Python - we're TypeScript zealots)
• Package Manager: Bun (NOT npm/yarn/pnpm)
• Format: Markdown (NOT HTML for basic content)
• Testing: Vitest when needed
• Voice: ElevenLabs TTS integration

---

🔐 Security

IMPORTANT: This is a PUBLIC template repository with sanitized examples.

DO NOT commit:

• API keys or secrets
• Personal email addresses or contact information
• Private repository references
• Any sensitive personal data

See SECURITY.md for complete security protocols.

---

💬 Community

• GitHub Issues: Report bugs or request features
• Discussions: Ask questions and share ideas
• Video: Watch PAI Overview

---

📝 Changelog

v0.9.0 (2025-12-01) - Platform Agnostic Release

This release focuses on making PAI fully portable and fork-friendly. Your AI, your identity, your system.

Observability Dashboard

• Complete real-time agent monitoring at .claude/Observability/
• WebSocket streaming of all agent activity
• Live pulse charts, event timelines, and swim lanes
• Multiple themes (Tokyo Night, Nord, Catppuccin, etc.)
• Security obfuscation for sensitive data
• Run with: ~/.claude/Observability/manage.sh start

Genericized Agent Identity

• All agent references now use process.env.DA || 'main'
• No more hardcoded names - your DA name flows through the entire system
• Observability dashboard shows your configured identity
• Color system supports any DA name with intelligent defaults

Platform-Agnostic Configuration

• Added _envDocs to settings.json explaining all environment variables
• Clear separation: settings.json for identity/paths, .env for API keys
• DA (Digital Assistant name) - your AI's identity
• PAI_DIR - root directory for all configuration
• TIME_ZONE - configurable timezone for timestamps

Statusline Enhancements

• Terminal color compatibility mode for different terminal emulators
• Dynamic CC version display
• Cleaner greeting format

Skill System Improvements

• Canonical TitleCase file naming throughout
• Standardized skill-workflow-notification script for dashboard detection
• All paths use ${PAI_DIR}/ for location-agnostic installation

Previous Releases

v0.7.0 - Research & Documentation

• Comprehensive research skill with 10 workflows
• Path standardization (${PAI_DIR}/)
• PAI_CONTRACT.md defining core guarantees
• Self-test validation system

v0.6.0 and earlier

• BrightData four-tier progressive web scraping
• Art and story-explanation skills
• Voice server with ElevenLabs integration
• Hook system architecture

For detailed changelogs, see the CHANGELOG-*.md files in the repository root.

---

📜 License

MIT License - see LICENSE file for details.

---

🙏 Acknowledgments

Built on Claude Code by Anthropic.

Inspired by the idea that AI systems need scaffolding to be reliable. This is that scaffolding.

---

Start clean. Start small. Build the AI infrastructure you need.

⬆ Back to Top