Formerly Clawdbot & Moltbot

What is OpenClaw?

OpenClaw (also known as Open Claw) is an open-source AI agent platform that runs on your own infrastructure and connects to the messaging channels your team already uses.

133K+GitHub Stars
700+Skills on ClawHub
15+Channels

What OpenClaw Does

OpenClaw turns any LLM (Claude, GPT-4, Gemini, or open-source models) into an interactive AI agent that lives inside your messaging apps. Instead of switching to a separate chat interface, you talk to your AI agent through WhatsApp, Slack, Discord, or any other connected channel.

The platform is fully self-hosted, meaning your conversations, data, and configurations stay on your own servers. You control which LLM provider to use, what skills (plugins) are installed, and how the agent behaves through custom skill development and personality configuration.

Self-Hosted

Runs on your own infrastructure. Your data never leaves your servers unless you choose to connect external services.

Multi-Channel

Connects to WhatsApp, Telegram, Slack, Discord, Microsoft Teams, Signal, and iMessage from a single instance.

Extensible

Over 700 community skills available on ClawHub, plus the ability to build custom skills using the skill.md standard.

Customizable

Define your AI agent's personality, guardrails, and brand voice through the soul.md configuration file.

Clawdbot → Moltbot → OpenClaw

Clawdbot

Original name

The project launched as Clawdbot — an open-source AI agent that could run locally and connect to messaging platforms.

Moltbot

First rename

The project was renamed to Moltbot with a restructured codebase, new plugin system, and improved configuration format.

OpenClaw

Current name

The project settled on OpenClaw as its final name. It now has 133K+ GitHub stars and a thriving community.

How OpenClaw Works

Every interaction follows the same request flow: message in, AI processing, response out.

1

Message Received

A user sends a message through a connected channel (WhatsApp, Slack, Discord, etc.).

2

Gateway Routes

The Gateway server receives the request, authenticates it, applies rate limiting, and routes it to the AI backend.

3

LLM Processes

The message is sent to the configured LLM (Claude, GPT-4, Gemini, or another provider) along with the soul.md personality and relevant skills.

4

Response Delivered

The AI’s response is sent back through the same channel, with any skill outputs (web searches, API calls, file operations) included.

Core Components

Gateway

The HTTP server that handles incoming requests, routes them to the AI backend, and enforces authentication and rate limiting.

ClawHub

Community marketplace with 700+ skills (plugins). Skills are community-submitted and range from web scrapers to calendar integrations.

skill.md

The standard file format for defining OpenClaw skills. It specifies what a skill does, its inputs/outputs, and execution logic.

soul.md

Personality configuration file that defines how the AI speaks, what topics it avoids, brand guidelines, and behavioral guardrails.

Common Challenges

Security Risks

Out-of-the-box OpenClaw has known risks: exposed gateways, unvetted ClawHub skills, and prompt injection vulnerabilities. Proper configuration is essential.

Cost Management

LLM API costs can escalate quickly with heavy usage, compaction events, and runaway automations. Smart model routing and token budgets help control spending.

Complexity

Setting up integrations, managing configs, and debugging issues requires technical knowledge. Deployment guides and expert help reduce the learning curve.

Getting Started

New to OpenClaw?

Start with the fundamentals and learn your way around the platform.

Ready to Deploy?

Pick your platform and follow one of the step-by-step deployment guides.

Need Expert Help?

Connect with vetted OpenClaw specialists for setup, security, and optimization.

Frequently Asked Questions

Need Help with OpenClaw?

OpenClaw Experts connects you with vetted specialists for setup, security, cost optimization, and custom development.