OpenClaw Has 385,000 Stars — and Sandboxing Off by Default

OpenClaw is a highly popular open-source personal AI assistant with 385,400 GitHub stars and 81,018 forks. It integrates with multiple messaging platforms including WhatsApp, Telegram, Slack, Discord, and Signal. A critical security concern is that sandboxing is disabled by default, which could pose risks if the assistant executes untrusted code. The project demonstrates significant community adoption but highlights the importance of secure default configurations in AI agent frameworks.

Key Points

• •OpenClaw has achieved massive adoption with 385,400 stars and 81,018 forks on GitHub
• •Multi-platform integration: supports WhatsApp, Telegram, Slack, Discord, and Signal for AI interactions
• •Sandboxing is disabled by default, creating potential security vulnerabilities for code execution
• •Functions as a personal AI assistant framework for developers and end-users
• •High community engagement suggests strong demand for accessible AI agent platforms
• •Security-first approach requires enabling sandboxing manually for safe code execution
• •Suitable for building conversational AI agents across distributed communication channels

Workflow Diagram

Concepts