An advanced security auditor designed to protect your environment by scanning Openclaw Skills for malicious code, prompt injections, and supply chain vulnerabilities.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install benlee-skillguard
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install benlee-skillguard 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 Skill Guard?
Skill Guard is a comprehensive security scanner built specifically for the Openclaw ecosystem. It provides a robust defensive layer by automatically auditing Openclaw Skills for a wide array of threats, including shell injections, data exfiltration, and credential theft. By utilizing smart domain analysis and tamper detection, it ensures that your AI agents operate within a secure and trustworthy framework.
Designed with portability in mind, this tool relies solely on the Python 3 standard library, making it easy to deploy across various environments. It goes beyond simple pattern matching by analyzing domain reputations, decoding obfuscated payloads, and identifying suspicious file modifications, providing developers with the peace of mind needed when integrating third-party Openclaw Skills.
Skill Guard Use Cases
- Auditing all installed Openclaw Skills for malicious code or backdoors.
- Verifying the integrity of newly downloaded skills before execution.
- Setting up automated cron jobs to monitor for unauthorized file changes in skill directories.
- Identifying hidden prompt injection attempts within skill documentation or source files.
- Scanning for hardcoded secrets like AWS keys, GitHub tokens, and browser credentials.
How Skill Guard Works
- The scanner initializes by targeting the Openclaw Skills directory or a specific path provided by the user.
- It performs a recursive static analysis of all files, including code, documentation, and configuration metadata.
- The engine evaluates the skill against a complex scoring system, penalizing suspicious patterns like unknown API calls or sensitive file access.
- A baseline of file hashes is established and compared during subsequent runs to identify any tampering or unexpected updates.
- The tool generates a final risk report, classifying the skill as Clean, Suspicious, or Malicious, complete with actionable remediation steps.
Skill Guard Setup
Skill Guard requires only Python 3 and has no external dependencies. To begin scanning your Openclaw Skills, use the following commands:
# Audit all installed skills in the default directory
python3 ~/clawd/skills/skill-guard/scripts/skillguard.py scan
# Check a specific skill directory for vulnerabilities
python3 ~/clawd/skills/skill-guard/scripts/skillguard.py check ~/clawd/skills/target-skill-name
# Generate a detailed Markdown report
python3 ~/clawd/skills/skill-guard/scripts/skillguard.py scan --report audit-results.md
Skill Guard Data Schema & Taxonomy
Skill Guard organizes its security metadata and audit logs to provide clear visibility into the state of your Openclaw Skills:
| Component | Purpose |
|---|---|
baselines.json |
Stores SHA-256 hashes of files to detect unauthorized modifications. |
origin.json |
Tracks the ClawHub source version to ensure supply chain integrity. |
| JSON Output | Machine-readable scan results for integration with other security tools. |
| Markdown Reports | Human-readable summaries containing risk scores and recommendations. |
Skill Guard Advanced Features
- Smart Domain Analysis that cross-references outbound requests against an allowlist of 80+ trusted APIs.
- Automated decoding of Base64 and Hex-encoded payloads to uncover hidden malicious scripts.
- Unicode homoglyph detection to prevent spoofing of filenames using lookalike characters.
- Persistence detection targeting modifications to crontab, launchd, or shell RC files.
- Comprehensive secret scanning for AWS, GitHub, OpenAI, and Stripe credentials.
SKILL.md
Loading
Related Openclaw Skills
Willhaben Marketplace Automator
An AI-driven browser automation skill for generating and publishing optimized listings on the Willhaben.at Austrian marketplace.
JSON2Video Pinterest
A powerful automation tool for creating professional, AI-powered vertical videos for Pinterest and social media.
Volcengine Configuration
A comprehensive onboarding and configuration guide to integrate Volcengine Ark coding models into your OpenClaw environment.
Azure OpenAI Proxy
A lightweight Node.js proxy that enables seamless integration between Azure OpenAI and OpenClaw by resolving API versioning and URL path conflicts.
YouTube Archiver
A powerful utility to archive YouTube playlists into local Markdown notes enhanced with transcripts, tags, and AI-generated summaries.
BenOS Memory Core
A high-performance runtime memory module for BenOS agents to manage session states, active initiatives, and decision logging.
Featured*
