📄 中文摘要
ClawMoat 是一个开源的运行时安全层，旨在保护计算机、凭证和敏感数据，免受恶意或不当 AI 代理行为的影响，尤其是那些基于 OpenClaw 平台构建的代理。它像一个“安全护城河”一样环绕在计算机周围，能够在危险行为到达系统之前进行阻挡，从而确保用户的安全和隐私。ClawMoat 通过实时监控和拦截潜在威胁，为用户提供了一种有效的防护机制，适用于各种使用场景。
ClawMoat — 保护您的计算机免受 AI 代理威胁
出处: ClawMoat — Protecting Your Machine from AI Agent Threats
发布: 2026年4月1日
📄 English Summary
ClawMoat — Protecting Your Machine from AI Agent Threats
ClawMoat is an open-source runtime security layer designed to protect computers, credentials, and sensitive data from malicious or careless actions by AI agents, particularly those built on platforms like OpenClaw. It functions as a 'security moat' around the machine, blocking dangerous behaviors before they can reach the system. By providing real-time monitoring and interception of potential threats, ClawMoat offers an effective protective mechanism for users, applicable in various scenarios.