# Crawsecure

> Offline security scanner for ClawHub skills — detect unsafe patterns before installation.

- Skill: `johnalbertini14-glitch/crawsecure` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add johnalbertini14-glitch/crawsecure`
- Raw SKILL.md: https://api.skillmd.com/api/skills/johnalbertini14-glitch/crawsecure/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- Author: johnalbertini14-glitch (https://skillmd.com/u/johnalbertini14-glitch)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/johnalbertini14-glitch/crawsecure

---

Offline security scanner for ClawHub skills — detect unsafe patterns before installation.

# CrawSecure

CrawSecure is an **offline security analysis skill** designed to help users evaluate potential risks in ClawHub / OpenClaw skills **before installing or trusting them**.

It promotes safer usage, transparency, and awareness when working with third-party skills.

---

## 🔍 What CrawSecure does

- Analyzes skill-related files locally
- Detects potentially dangerous patterns
- Highlights security risks clearly
- Helps users make informed decisions before installation

---

## 🚨 Risk Signals Analyzed

- Dangerous command patterns (e.g. destructive or execution-related behavior)
- References to sensitive files or credentials
- Indicators of unsafe or risky practices

---

## 🔒 Security Philosophy

- Read-only analysis
- No network access
- No code execution
- No file modifications

CrawSecure exists to **increase trust** inside the ClawHub ecosystem.

---

## ⚙️ Execution Model

CrawSecure does **NOT** execute or install any third-party code.

This skill provides a **local CLI tool** that users run manually.

### Using npx (recommended)

```bash
npx crawsecure ./path-to-skill