Ora

Omniscient. Real-time. Autonomous.

The framework for autonomous AI development with unified memory across all your tools.

---

What is Ora?

Ora synchronizes Claude, Cursor, GitHub, Pinecone, LangSmith, and Slack into a single autonomous development system with omniscient memory. One heartbeat. Perfect sync.

Ora enables you to create isolated, autonomous agent systems for any project in under 5 minutes. Each project gets its own PostgreSQL schema, Pinecone namespace, Slack channel, and agent workforce—completely isolated from other projects.

---

Key Features

🧠 Omniscient Memory

• Pinecone Vector Database: Maintains perfect context across all agents and conversations
• PostgreSQL State: Persistent agent state, decision history, and performance metrics
• Unified Knowledge: All project knowledge accessible to every agent instantly

⚡ 15-Minute Oversight

• Daily Strategic Reviews: High-level oversight without micromanagement
• Autonomous Execution: Agents work independently between reviews
• Perfect Coordination: File-based communication ensures reliable agent coordination

🚀 Rapid Project Setup

• 5-Minute Bootstrap: ora-init my-project creates everything you need
• Complete Isolation: Each project gets isolated infrastructure
• Zero Configuration: Works out of the box with sensible defaults

🤖 Agent Templates

• Pre-built Archetypes: Research, Implementation, QA, Documentation, Infrastructure agents
• Instant Deployment: Deploy agents in seconds using templates
• Proven Patterns: Based on production MCP-WP system (12+ agents, 2,426 vectors)

🔒 Perfect Isolation

• PostgreSQL: Schema isolation (ora_{{PROJECT_NAME}})
• Pinecone: Namespace isolation ({{PROJECT_NAME}})
• Slack: Channel isolation (#ora-{{PROJECT_NAME}}-agents)
• File System: Separate project directories

---

Quick Start

Prerequisites

• PostgreSQL database (local or remote)
• Pinecone account and API key
• Slack workspace and bot token
• Node.js and Python 3
• Git

Installation

# Clone the repository
git clone https://github.com/jeffmichaeljohnson-tech/ora.git
cd ora

# Set up environment variables
export DB_HOST=localhost
export DB_PORT=5432
export DB_NAME=postgres
export DB_USER=postgres
export DB_PASSWORD=your-password
export PINECONE_API_KEY=your-pinecone-key
export PINECONE_INDEX_NAME=ora-framework-index
export SLACK_BOT_TOKEN=your-slack-token

Create Your First Project

# Initialize a new project
cd infrastructure/bootstrap
./ora-init.sh my-awesome-project

# Your autonomous workforce is ready in < 5 minutes!

That's it! Ora will:

1. ✅ Create PostgreSQL schema (ora_my-awesome-project)
2. ✅ Create Pinecone namespace (my-awesome-project)
3. ✅ Create Slack channel (#ora-my-awesome-project-agents)
4. 📋 Guide you through LangSmith setup (manual)
5. 📋 Guide you through GitHub setup (manual)

Deploy Your First Agents

# Deploy agents using templates
cd agent-templates/deploy
./ora-hire-agent.sh research-agent my-awesome-project
./ora-hire-agent.sh implementation-agent my-awesome-project
./ora-hire-agent.sh qa-agent my-awesome-project

---

Architecture Overview

Ora is built on four core principles:

1. Omniscient Memory

Every agent has access to the complete project knowledge base through Pinecone vectors and PostgreSQL state. No context is lost.

2. Infrastructure Isolation

Each project gets completely isolated infrastructure:

• PostgreSQL Schema: ora_{{PROJECT_NAME}} - separate tables, no data mixing
• Pinecone Namespace: {{PROJECT_NAME}} - vector isolation via namespaces
• Slack Channel: #ora-{{PROJECT_NAME}}-agents - dedicated communication channel
• File System: /projects/{{PROJECT_NAME}}/ - separate directories

3. File-Based Communication

Agents communicate via JSON files in inbox/ and outbox/ directories. This ensures:

• Reliability (no network dependencies)
• Auditability (all messages persisted)
• Simplicity (easy to debug and monitor)

4. 15-Minute Oversight

The Director (Claude) reviews progress every 15 minutes, making strategic decisions while agents execute autonomously.

---

Project Structure

ora/
├── infrastructure/          # Infrastructure templates
│   ├── postgres/           # PostgreSQL schema templates
│   ├── pinecone/           # Pinecone namespace setup
│   ├── slack/              # Slack channel creation
│   ├── langsmith/          # LangSmith project setup
│   ├── github/             # GitHub repository templates
│   └── bootstrap/          # Master bootstrap script (ora-init.sh)
├── agent-templates/        # Agent archetype templates
│   ├── archetypes/         # Agent type templates
│   ├── schemas/            # Communication schemas
│   └── deploy/             # Agent deployment scripts
├── cli/                    # Command-line tools
├── docs/                   # Documentation
└── projects/               # Project instances

---

Documentation

• QUICKSTART.md - Get started in 10 minutes
• ARCHITECTURE.md - Deep dive into system design
• TROUBLESHOOTING.md - Common issues and solutions
• AGENT-TEMPLATES.md - Agent template guide
• TEST-RESULTS.md - System test results

---

Status

Built and proven in production — extracted from a live autonomous system that coordinated 27 AI agents over 5 months of continuous daily operation.

---

Contributing

Contributions welcome. Open an issue to discuss.

---

License

MIT License — see LICENSE

---

Support

• Documentation: See /docs directory
• Issues: GitHub Issues

---

Acknowledgments

Ora was extracted from a production autonomous system (MCP-WP) that coordinated 27 AI agents with 2,442 knowledge vectors over 5 months of continuous daily operation.

---

Built by Jeff Michael Johnson