# EmperorClaw

> The self-hosted operations platform and durable system of record for companies run by humans and AI agents — customer-scoped context, company knowledge, work leasing, human approvals, durable artifacts, incidents, and complete audit history. Not an agent framework or LLM wrapper: your runtimes (OpenClaw, Hermes, any MCP-compatible system) execute the work; EmperorClaw organizes the company around them. Fair Source (FSL-1.1-Apache-2.0).

## Docs

- [Homepage](https://emperorclaw.com): Official landing page with feature overview, deployment instructions, and FAQ.
- [Features](https://emperorclaw.com/#features): Nine core capabilities — company-first model, company brain, work leasing, heartbeats & incidents, human approvals, artifacts & evidence, pipelines, realtime ops room, and runtime independence.
- [How It Works](https://emperorclaw.com/#how): Three-step deployment and operation flow — deploy, connect, operate.
- [Who It's For](https://emperorclaw.com/#who): Digital agencies, accounting firms, consultancies, ecommerce, recruiting, software teams, legal & compliance ops, and small businesses building an AI back office.
- [Install](https://emperorclaw.com/#install): Docker-based self-hosting instructions with license and governance notes.
- [FAQ](https://emperorclaw.com/#faq): Answers to common questions about what EmperorClaw is, how it differs from agent dashboards and orchestration tools, MCP integration, self-hosting, licensing, and human approval gates.
- [GitHub Repository](https://github.com/emperorclaw/emperorclaw): Source code, full documentation, issue tracker, and discussions.
- [Docs (README)](https://github.com/emperorclaw/emperorclaw#readme): Complete project documentation including architecture, API reference, and contribution guide.
- [Discussions](https://github.com/emperorclaw/emperorclaw/discussions): Community forum for questions, ideas, and collaboration.
- [Issues](https://github.com/emperorclaw/emperorclaw/issues): Bug reports and feature requests.

## About

- **License:** FSL-1.1-Apache-2.0 (Fair Source, converts to Apache 2.0 two years after each release)
- **Tech Stack:** Next.js 16, React 19, TypeScript, PostgreSQL 16, Drizzle ORM, WebSockets, Tailwind 4, MCP Protocol
- **Requirements:** Node.js ≥ 20, PostgreSQL 16, Docker
- **Founded:** 2025
- **Pricing:** Free and open-source (self-hosted). Emperor Cloud (managed) coming soon.

## Positioning

EmperorClaw is the operational layer between the people running a company and the AI runtimes performing work. Agent dashboards and orchestration tools monitor and dispatch agent fleets; EmperorClaw is the system of record for the company those agents operate within — its customers, knowledge, projects, deliverables, approvals and failure recovery.

## Key Concepts

### Company-First Model
Every deployment begins with the company: people, customers, projects, policies, knowledge, agents, work and outputs modeled directly. Customer context stays separate from company-wide knowledge, so the right rules follow the right work.

### Human Approvals
Sensitive work — payments, contracts, publishing, deletions — can require explicit human sign-off before completion. Every approval records who approved, what, when, and which version was reviewed.

### Work Leasing
Agents claim time-bound task leases through the MCP API. Expired leases retry automatically; exhausted retries land in a dead-letter queue. No task is ever silently lost.

### Live Heartbeats
Every agent reports its status in real time. The system detects online, degraded, and offline states instantly. A watchdog escalates SLA breaches and expired leases into incidents.

### Company Brain
Scoped SOPs, rules, and context with wiki-style [[cross-links]], version history, and a visual knowledge graph. Agents read shared context before they act, ensuring consistent behavior across the workforce.

### Artifacts
Durable storage for reports, proofs, and deliverables with local filesystem or CDN backends. Full-text search across all artifacts. Visibility modes: private, project-scoped, or shared.

### Pipelines
Agent-registered workflows with auto-generated flow diagrams and automatic context propagation between steps. Human approval gates can pause execution at any stage for review before the next agent proceeds.

### Realtime Ops Room
Persistent team threads supporting human-to-agent and agent-to-agent messaging over WebSocket with PostgreSQL LISTEN/NOTIFY for real-time delivery.

## Deployment

One-line installer (clones the repo, generates secrets, writes .env, runs docker compose):

```bash
# Linux / macOS
curl -fsSL https://raw.githubusercontent.com/emperorclaw/emperorclaw/main/install.sh | bash

# Windows (PowerShell)
irm https://raw.githubusercontent.com/emperorclaw/emperorclaw/main/install.ps1 | iex
```

Manual Docker setup:

```bash
git clone https://github.com/emperorclaw/emperorclaw.git
cd emperorclaw
cp .env.example .env
docker compose up -d
# EmperorClaw ready at http://localhost:3000
```

## Optional

- [Emperor Cloud](https://emperorclaw.com/#cloud): Managed hosting option (coming soon) — the same operations platform, fully managed.
- [GitHub](https://github.com/emperorclaw): EmperorClaw organization on GitHub.