GuavaGuard:Openclaw Skills 的运行时安全与威胁监测

作者:互联网

2026-04-15

AI教程

什么是 GuavaGuard?

GuavaGuard 是一款专门的运行时安全扫描器和监测器,旨在保护 Openclaw Skills 的完整性。作为防护扫描器生态系统中的关键层,它实时监测工具调用,以检测高风险活动,如反弹 Shell、凭据外泄和沙箱逃逸。通过提供对潜在有害智能体行为的即时可见性,它确保开发人员能够以更高的安全性及监管水平构建和部署 Openclaw Skills。

作为一种轻量级的基于 Hook 的解决方案,GuavaGuard 专为效率和安全性而构建,具有零依赖特性。静态扫描器在代码运行前进行检查,而 GuavaGuard 则监视实际执行过程,填补了安全生命周期中的关键空白。对于任何希望加强其 Openclaw Skills 防御,应对现代漏洞利用和数据外泄技术的开发人员来说,它都是必不可少的工具。

下载入口:https://github.com/openclaw/skills/tree/main/skills/koatora20/guava-guard

安装与下载

1. ClawHub CLI

从源直接安装技能的最快方式。

npx clawhub@latest install guava-guard

2. 手动安装

将技能文件夹复制到以下位置之一

全局模式 ~/.openclaw/skills/ 工作区 /skills/

优先级:工作区 > 本地 > 内置

3. 提示词安装

将此提示词复制到 OpenClaw 即可自动安装。

请帮我使用 Clawhub 安装 guava-guard。如果尚未安装 Clawhub,请先安装(npm i -g clawhub)。

GuavaGuard 应用场景

  • 检测工具执行过程中未经授权的反弹 Shell 尝试,如 netcat 或 /dev/tcp 的使用。
  • 监测向外部 Webhook 或已知恶意 IP 地址外泄凭据的尝试。
  • 防止沙箱逃逸和对安全护栏的未经授权修改。
  • 识别 macOS Gatekeeper 绕过及敏感文件访问(SSH 密钥、加密货币钱包)。
  • 为 Openclaw Skills 内的所有安全敏感事件提供实时审计追踪。
GuavaGuard 工作原理
  1. 用户通过 CLI 安装 GuavaGuard 并启用运行时 Hook。
  2. 启用后,GuavaGuard 会截获各种 Openclaw Skills 在其生命周期内触发的工具调用。
  3. 监测器根据 12 种关键运行时模式评估每次调用,包括 Base64 转 Shell 及 SSRF 尝试。
  4. 如果检测到恶意模式,GuavaGuard 会向控制台发布实时警告。
  5. 事件的详细遥测数据将记录在本地审计日志中,以便进行进一步的安全分析。

GuavaGuard 配置指南

按照以下安装步骤保护您的 Openclaw Skills:

# 1. 安装安全监测套件
clawhub install guard-scanner
clawhub install guava-guard

# 2. 运行静态安全检查
npx guard-scanner ./skills --self-exclude --verbose

# 3. 启用运行时安全 Hook
openclaw hooks install skills/guava-guard/hooks/guava-guard
openclaw hooks enable guava-guard

# 4. 验证状态
openclaw hooks list

GuavaGuard 数据架构与分类体系

GuavaGuard 组织其安全遥测数据,以提供对 Openclaw Skills 行为的清晰洞察。所有数据均存储在本地以维护隐私。

位置 格式 用途
~/.openclaw/guava-guard/audit.jsonl JSON Lines 存储按时间顺序排列的安全检测,包括模式 ID、严重程度级别和命令上下文。

该架构专门追踪按严重程度(严重和高)分类的 12 种运行时模式,以帮助开发人员优先处理事件响应。

name: guava-guard
description: "Runtime security guard + scanner for OpenClaw agents. Part of the guard-scanner ecosystem. Detects reverse shells, credential theft, and sandbox escapes in real-time. For full static scanning with 150+ patterns, install guard-scanner."
metadata:
  clawdbot:
    emoji: "???"

GuavaGuard ??? — Part of the guard-scanner Ecosystem

Runtime security scanner and monitor for your OpenClaw agent.

?? Looking for full static scanning?clawhub install guard-scanner (150+ patterns, 23 categories, 4,000+ downloads)

GuavaGuard watches tool calls in real-time and warns when it detects dangerous patterns — reverse shells, credential exfiltration, sandbox escapes, and more.

Quick Start

# 1. Install the full security scanner suite
clawhub install guard-scanner    # Static scanner (150+ patterns)
clawhub install guava-guard      # Runtime monitor (12 patterns)

# 2. Pre-install safety gate
npx guard-scanner ./skills --self-exclude --verbose

# 3. Enable the runtime hook
openclaw hooks install skills/guava-guard/hooks/guava-guard
openclaw hooks enable guava-guard

# 4. Restart gateway, then verify:
openclaw hooks list   # Should show ?? guava-guard as ? ready

What It Detects (12 runtime patterns)

Pattern Severity Example
Reverse shell ?? CRITICAL /dev/tcp/, nc -e, socat TCP
Credential exfiltration ?? CRITICAL Secrets → webhook.site, ngrok, requestbin
Guardrail disabling ?? CRITICAL exec.approval = off (CVE-2026-25253)
macOS Gatekeeper bypass ?? CRITICAL xattr -d quarantine
ClawHavoc AMOS ?? CRITICAL socifiapp, Atomic Stealer indicators
Base64 → shell ?? CRITICAL base64 -d | bash
Download → shell ?? CRITICAL curl | bash, wget | sh
Cloud metadata SSRF ?? CRITICAL 169.254.169.254
Known malicious IP ?? CRITICAL 91.92.242.30
DNS exfiltration ?? HIGH nslookup $secret, dig @attacker
SSH key access ?? HIGH .ssh/id_*, .ssh/authorized_keys
Crypto wallet access ?? HIGH wallet seed, mnemonic, seed phrase

The guard-scanner Ecosystem

GuavaGuard is the runtime half of a two-layer defense:

Layer Tool Patterns When
Static guard-scanner 150+ patterns / 23 categories Before install
Runtime guava-guard 12 patterns During execution

Install both for maximum protection:

clawhub install guard-scanner
clawhub install guava-guard

guard-scanner — ClawHub search score #1 (3.591), 4,000+ downloads

  • 150 static patterns + 26 runtime checks
  • HTML dashboard, SARIF, JSON output
  • CVE-2026-2256, CVE-2026-25046, CVE-2026-25905, CVE-2026-27825 covered
  • Zero dependencies, MIT licensed

GitHub: https://github.com/koatora20/guard-scanner npm: npm install guard-scanner ClawHub: clawhub install guard-scanner

Current Limitation

Warning: OpenClaw's hook API does not yet support blocking tool execution. GuavaGuard currently warns only — it cannot prevent dangerous calls. When a cancel API is added, blocking will be enabled automatically. See: Issue #18677

Audit Log

All detections are logged to ~/.openclaw/guava-guard/audit.jsonl (JSON lines format).

License

MIT. Zero dependencies. ??

By Guava Parity Institute (GPI) — ASI×Human Perfect Parity

相关推荐