skill-guard
Scan ClawHub skills for security vulnerabilities BEFORE installing. Use when installing new skills from ClawHub to detect prompt injections, malware payloads, hardcoded secrets, and other threats. Wraps clawhub install with Snyk Agent Scan pre-flight checks.
Why use this skill?
Use skill-guard to scan ClawHub skills for prompt injections, malware, and secrets before installation. Ensure your agent remains secure with Snyk Agent Scan.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/firefrog-pepe/skill-guard-snyk-agent-scanWhat This Skill Does
skill-guard acts as the critical security perimeter for your OpenClaw agent ecosystem. It serves as a pre-install gatekeeper that verifies the integrity and safety of any skill sourced from ClawHub. Rather than blindly trusting remote code, skill-guard downloads candidate skills into a segregated staging environment in /tmp/. It then deploys the Snyk Agent Scan engine to perform a deep inspection, hunting for prompt injections, hardcoded secrets, data exfiltration vectors, and malicious command chains. By moving security to the client-side, it bridges the gap that standard server-side antivirus scanners miss, ensuring your agent environment remains free of toxic instructions or hidden exfiltration payloads.
Installation
Installation is handled via a secure wrapper script that replaces the standard clawhub install command. To integrate skill-guard, ensure you have the SNYK_TOKEN configured in your environment variables for the scanner to function. Simply replace your standard installation command with the provided safe-install utility:
./scripts/safe-install.sh <skill-slug>
This script automatically handles the staging, scanning, and installation process, providing exit codes to let you know the health of the skill. Always run this script whenever you plan to expand your agent's capabilities with new third-party skills.
Use Cases
Use skill-guard whenever you are integrating new third-party skills, especially those from untrusted or community-driven repositories. It is essential for enterprise security compliance, as it prevents "Shadow AI" behaviors where skills might attempt to exfiltrate private conversation logs or local file data to unknown external URLs. It is also a requirement for developers auditing their own local projects to ensure that internal dependencies do not accidentally contain security regressions or exposed API keys.
Example Prompts
- "Safe-install the 'data-summarizer' skill from the public hub and verify its security posture before I give it access to my documents."
- "Check if the new sentiment analysis skill I downloaded has any hardcoded credentials or data exfiltration URLs using the skill-guard scanner."
- "Attempt to install the 'gpt-finance-helper' via safe-install, and tell me if it triggers any prompt injection warnings during the Snyk analysis."
Tips & Limitations
To get the most out of skill-guard, always keep your Snyk Agent Scan definitions updated. Note that this tool is designed for pre-install validation; it does not monitor the skill's behavior during runtime once installed. If a scan returns exit code 3, check your SNYK_TOKEN and network configuration immediately, as the agent is running unprotected during this period. Do not bypass the safe-install.sh workflow; manual installation via clawhub directly circumvents all protective barriers provided by this skill.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-firefrog-pepe-skill-guard-snyk-agent-scan": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-read, file-write, code-execution