skill-guard

Scan staged skills for prompt injections, data exfiltration, and hardcoded secrets.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/kisslex2013-alt/sivraJ-OpenClaw --skill skill-guard
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-guard
Source: https://github.com/kisslex2013-alt/sivraJ-OpenClaw/tree/main/skills/skill-guard
Command: npx skills add https://github.com/kisslex2013-alt/sivraJ-OpenClaw --skill skill-guard

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires clawhub, uvx, and includes scripts (resource) components.

What problem does it solve?

Skills in ClawHub can introduce security risks before installation; skill-guard acts as a pre-install gate that stages, scans, and blocks unsafe content to prevent compromised skills from entering your workspace.

Core Features & Use Cases

  • Pre-install scanning of staged skills with mcp-scan to detect prompt injections, data exfiltration attempts, and hardcoded secrets.
  • Safe staging and controlled installation: skills are downloaded to a temporary area and only moved into place after a clean scan.
  • Quarantine and review workflow: suspicious results are set aside for manual review before any installation.

Quick Start

Use the secure-install script to stage a skill, run a pre-install scan, and install only if no threats are detected.

Frequently Asked Questions about skill-guard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I scan ClawHub skills for security risks before installation?▼

Pre-install security scanning for ClawHub skills works by downloading the skill to a dedicated staging area, running mcp-scan tooling to detect prompt injections, data exfiltration, and hardcoded secrets, and only moving the files into your workspace if no threats are found.

What security vulnerabilities should I check for when installing MCP skills?▼

When installing MCP skills, you should check for prompt injections, data exfiltration attempts, and hardcoded secrets. A pre-install scanning gate isolates the skill files in a staging directory and blocks installation if it detects any of these security threats.

Do I need clawhub and uvx installed to run pre-install skill scans?▼

Yes, you need the clawhub CLI and uvx installed to run pre-install skill scans. The scanning workflow requires the clawhub CLI to manage the skill installation process and the mcp-scan tooling to perform the actual security checks during staging.

What happens to unsafe skills detected during the pre-install security scan?▼

Unsafe skills detected during the pre-install security scan are moved to a quarantine area for manual review. This quarantine workflow ensures that suspicious code is set aside and prevented from entering your workspace until you complete a manual security review.

Can I review quarantined skills before allowing them into my workspace?▼

Yes, you can review quarantined skills before allowing them into your workspace. The pre-install security gate sets aside suspicious scan results into a quarantine area, enabling a controlled manual review workflow before any files are moved into the final installation directory.