Shield CN (Security Guard) for Openclaw

A specialized security skill designed to protect AI agents from prompt injection, data leaks, and credential theft within Chinese digital ecosystems.

onlyloveher
v2284.0.0
Mar 27, 2026
0
1.5k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install shield-cn

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install shield-cn using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Shield CN (Security Guard)?

Shield CN is a robust security enhancement for AI agents, specifically optimized for the nuances of the Chinese language and popular domestic platforms. Unlike generic tools, it provides localized protection against prompt hijacking and credential leaks on services like Alibaba Cloud, Tencent Cloud, and WeChat. It acts as a lightweight, real-time firewall that ensures your Openclaw Skills operate safely in production or development environments by filtering malicious inputs and restricting access to sensitive system files.

Shield CN (Security Guard) Use Cases

  • Defending against role-hijacking and malicious prompt injections written in Chinese.
  • Preventing unauthorized access to cloud credentials for Aliyun, Tencent, Huawei, and Baidu.
  • Monitoring and blocking sensitive data leaks to platforms like DingTalk, Feishu, and WeChat.
  • Generating detailed security audit reports to identify vulnerabilities in memory files or environment settings.

How Shield CN (Security Guard) Works

  1. The user installs the skill and configures the monitoring mode to either monitor, block, or audit.
  2. Shield CN runs a real-time guard script that intercepts agent interactions to detect suspicious patterns.
  3. It analyzes inputs using localized regex patterns for prompt injection and phishing URLs specific to Chinese services.
  4. The system enforces strict file access policies to protect .env files, private keys, and cloud configuration files.
  5. It generates automated security reports with actionable suggestions and risk scores to maintain high security standards for Openclaw Skills.

Shield CN (Security Guard) Setup

Install the skill using the CLI:

skillhub install shield-cn

Configure your protection rules in shield-config.json and start the guard:

cd skills/shield-cn
python3 scripts/shield-guard.py --config shield-config.json

Shield CN (Security Guard) Data Schema & Taxonomy

Component Description
shield-config.json Configuration file defining security modes, log levels, and blocked keywords.
security-report-*.md Markdown reports containing security scores and audit findings.
HIJACK_PATTERNS A library of regex patterns specifically tuned for Chinese prompt injection attempts.
PHISHING_DOMAINS A maintained blacklist of domains used for WeChat, DingTalk, and Aliyun phishing.

Shield CN (Security Guard) Advanced Features

  • Context-aware analysis to identify fragmented attacks spread across multiple conversation turns.
  • Real-time alert integration with DingTalk and Feishu webhooks for instant notifications.
  • Specialized credential protection for WeChat mini-program AppIDs and secrets.
  • Lightweight execution mode that operates independently without requiring modifications to core agent files in Openclaw Skills.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*