by yucai233 · Agent Tool · ★ 20
AgentUAC A simple tool to private agent from dangerous operations AgentUAC (User Account Control for AI) 为本地 AI 智能体提供运行时安全拦截与审批层。 概述 AgentUAC 是一个用于拦截 AI 智能体(如 OpenClaw, AutoGPT, LangChain 应用)高危操作的运行时安全工具。当 AI 尝试执行危险操作(如读取敏感文件、执行删除命令)时,系统会暂停执行并等待用户审批。 核心特性 零侵入: 通过 Node.js 机制自动挂载 同步阻塞: 使用 + 实现真正的线程阻塞 跨框架兼容: 通过 Module require hook 拦截,不依赖特定框架 Worker 线程分离: 审批逻辑在独立线程中运行,不阻塞主应用 危险操作拦截 文件系统 (fs) 读取敏感文件: , , 写入敏感路径 删除文件/目录 命令执行 (childprocess) , , , , , 项目结构 安装 使用方法 方式一: CLI 包装器 (推荐) 方式二: 直接 require 配置 默认危险路径和命令定义在 s
| Stars | 20 |
| Forks | 12 |
| Language | TypeScript |
| Category | Agent Tool |
| License | MIT |
| Quality Score | 63.1250278108736/100 |
| Last Updated | 2026-05-23 |
| Created | 2026-03-16 |
| Platforms | node |
| Est. Tokens | ~2k |
These tools work well together with AgentUAC for enhanced workflows:
Explore other popular agent tool tools:
AgentUAC is A simple tool to private agent from dangerous operations. It is categorized as a Agent Tool with 20 GitHub stars.
AgentUAC is primarily written in TypeScript.
You can find installation instructions and usage details in the AgentUAC GitHub repository at github.com/yucai233/AgentUAC. The project has 20 stars and 12 forks, indicating an active community.
AgentUAC is released under the MIT license, making it free to use and modify according to the license terms.